- Computers & electronics
- Computer components
- System components
- Motherboards
- NXP Semiconductors
- ColdFire MCF52233
- Reference manual
advertisement
MCF52235 ColdFire
®
Integrated
Microcontroller Reference Manual
Devices Supported:
MCF52230
MCF52231
MCF52233
MCF52234
MCF52235
Document Number:
MCF52235RM
Rev. 0
04/2006
How to Reach Us:
Home Page: www.freescale.com
E-mail: [email protected]
USA/Europe or Locations Not Listed:
Freescale Semiconductor
Technical Information Center, CH370
1300 N. Alma School Road
Chandler, Arizona 85224
+1-800-521-6274 or +1-480-768-2130 [email protected]
Europe, Middle East, and Africa:
Freescale Halbleiter Deutschland GmbH
Technical Information Center
Schatzbogen 7
81829 Muenchen, Germany
+44 1296 380 456 (English)
+46 8 52200080 (English)
+49 89 92103 559 (German)
+33 1 69 35 48 48 (French) [email protected]
Japan:
Freescale Semiconductor Japan Ltd.
Headquarters
ARCO Tower 15F
1-8-1, Shimo-Meguro, Meguro-ku,
Tokyo 153-0064, Japan
0120 191014 or +81 3 5437 9125 [email protected]
Asia/Pacific:
Freescale Semiconductor Hong Kong Ltd.
Technical Information Center
2 Dai King Street
Tai Po Industrial Estate
Tai Po, N.T., Hong Kong
+800 26668334 [email protected]
For Literature Requests Only:
Freescale Semiconductor Literature Distribution Center
P.O. Box 5405
Denver, Colorado 80217
1-800-441-2447 or 303-675-2140
Fax: 303-675-2150
Information in this document is provided solely to enable system and software implementers to use Freescale Semiconductor products. There are no express or implied copyright licenses granted hereunder to design or fabricate any integrated circuits or integrated circuits based on the information in this document.
Freescale Semiconductor reserves the right to make changes without further notice to any products herein. Freescale Semiconductor makes no warranty, representation or guarantee regarding the suitability of its products for any particular purpose, nor does Freescale Semiconductor assume any liability arising out of the application or use of any product or circuit, and specifically disclaims any and all liability, including without limitation consequential or incidental damages. “Typical” parameters that may be provided in Freescale
Semiconductor data sheets and/or specifications can and do vary in different applications and actual performance may vary over time. All operating parameters, including “Typicals”, must be validated for each customer application by customer’s technical experts. Freescale Semiconductor does not convey any license under its patent rights nor the rights of others.
Freescale Semiconductor products are not designed, intended, or authorized for use as components in systems intended for surgical implant into the body, or other applications intended to support or sustain life, or for any other application in which the failure of the Freescale Semiconductor product could create a situation where personal injury or death may occur. Should Buyer purchase or use Freescale Semiconductor products for any such unintended or unauthorized application, Buyer shall indemnify and hold Freescale
Semiconductor and its officers, employees, subsidiaries, affiliates, and distributors harmless against all claims, costs, damages, and expenses, and reasonable attorney fees arising out of, directly or indirectly, any claim of personal injury or death associated with such unintended or unauthorized use, even if such claim alleges that Freescale Semiconductor was negligent regarding the design or manufacture of the part.
Freescale™ and the Freescale logo are trademarks of Freescale
Semiconductor, Inc. All other product or service names are the property of their respective owners.
© Freescale Semiconductor, Inc. 2006. All rights reserved.
Document Number: MCF52235RM
Rev. 0
04/2006
Table of Contents
Chapter 1
Overview
Chapter 2
Signal Descriptions
2
C I/O Signals ................................................................................................................................10
Chapter 3
ColdFire Core
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary iii
Chapter 4
Enhanced Multiply-Accumulate Unit (EMAC)
Chapter 5
Cryptographic Acceleration Unit
Chapter 6
Random Number Generator Accelerator (RNGA)
Chapter 7
Clock Module
Chapter 8
Real Time Clock
iv
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 9
Power Management
Chapter 10
Reset Controller Module
Chapter 11
Static RAM (SRAM)
Chapter 12
Chip Configuration Module (CCM)
Chapter 13
System Control Module (SCM)
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary v
Chapter 14
General Purpose I/O Module
Chapter 15
Interrupt Controller Module
Chapter 16
Edge Port Module (EPORT)
Chapter 17
ColdFire Flash Module (CFM)
Chapter 18
Fast Ethernet Controller (FEC)
Chapter 19
Ethernet Physical Transceiver (EPHY) Block Description
vi
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 20
DMA Controller Module
Chapter 21
EzPort
Chapter 22
Programmable Interrupt Timer Modules (PIT0–PIT1)
Chapter 23
General Purpose Timer Module (GPT)
Chapter 24
DMA Timers (DTIM0–DTIM3)
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary vii
Chapter 25
Queued Serial Peripheral Interface (QSPI)
Chapter 26
UART Modules
I
2
Chapter 27
C Interface
2
C System Configuration .................................................................................................................3
2
C Programming Examples ............................................................................................................12
Chapter 28
Analog-to-Digital Converter (ADC)
Chapter 29
Pulse Width Modulation (PWM) Module
Chapter 30
FlexCAN
viii
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 31
Debug Module
Chapter 32
IEEE 1149.1 Test Access Port (JTAG)
Appendix A
Register Memory Map Quick Reference
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary ix
x
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 1
Overview
This chapter provides an overview of the major features and functional components of the MCF52235 family of microcontrollers. The MCF52235 family is a highly integrated implementation of the ColdFire
® family of reduced instruction set computing (RISC) microcontrollers that also includes the MC52230,
MCF52231, MC52233 and MC52234 . The differences between these parts are summarized in Table 1-1 .
This document is written from the perspective of the MC52235 .The MC52235 represents a family of highly-integrated 32-bit microcontrollers based on the V2 ColdFire microarchitecture. Featuring up to 32
Kbytes of internal SRAM and 256 Kbytes of Flash memory, four 32-bit timers with DMA request capability, a 4-channel DMA controller, fast Ethernet, a CAN module, an I
2
C™ module, 3 UARTs and a queued SPI, the MC52235 family has been designed for general-purpose industrial control applications. n enhanced multiply-accumulate unit (EMAC) and divider providing 56Drystone 2.1 MIPS at a frequency up to 60MHz from internal Flash. On-chip modules include the following:
• V2 ColdFire core with enhanced multiply-accumulate unit (EMAC)
• Cryptographic Acceleration Unit (CAU)
• 32 Kbytes of internal SRAM
• 256 Kbytes of on-chip Flash memory
• Fast Ethernet Controller (FEC) with on-chip transceiver (ePHY)
• Three universal asynchronous receiver/transmitters (UARTs)
• Controller area network 2.0B (FlexCAN) module
• Inter-integrated circuit (I 2 C) bus controller
• 12-bit analog-to-digital converter (ADC)
• Queued serial peripheral interface (QSPI) module
• Four-channel, 32-bit direct memory access (DMA) controller
• Four-channel, 32-bit input capture/output compare timers with optional DMA support
• Two 16-bit periodic interrupt timers (PITs)
• Programmable software watchdog timer
• Two interrupt controllers, each capable of handling up to 63 interrupt sources (126 total)
These devices are ideal for cost-sensitive applications requiring significant control processing for connectivity, data buffering, and user interface, as well as signal processing in a variety of key markets such as security, imaging, networking, gaming, and medical. This leading package of integration and high performance allows fast time to market through easy code reuse and extensive third party tool support.
To locate any published errata or updates for this document, refer to the ColdFire products website at http://www.freescale.com/coldfire .
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
1-1
Overview
1.1
MCF52235 Family Configurations
Table 1. MCF52235 Family Configurations
Module 52230 52231 52233 52234 52235
ColdFire Version 2 Core with EMAC (Enhanced
Multiply-Accumulate Unit)
System Clock
Performance (Dhrystone 2.1 MIPS)
Flash / Static RAM (SRAM)
Interrupt Controllers (INTC0/INTC1)
Fast Analog-to-Digital Converter (ADC)
Random Number Generator and Crypto
Acceleration Unit (CAU)
FlexCAN 2.0B Module
Fast Ethernet Controller (FEC) with on-chip interface (EPHY)
Four-channel Direct-Memory Access (DMA)
Software Watchdog Timer (WDT)
Programmable Interrupt Timer
Four-Channel General Purpose Timer
32-bit DMA Timers
QSPI
UART(s)
I
2
C
Eight/Four-channel 8/16-bit PWM Timer
General Purpose I/O Module (GPIO)
Chip Configuration and Reset Controller Module
Background Debug Mode (BDM)
JTAG - IEEE 1149.1 Test Access Port
1 x
128/32 Kbytes 128/32 Kbytes x x
x x
x x x x x x x
3 x
4 x x
2
x x x x x x x x x
3 x
4 x x
2 x x x
60 MHz
56 x x x x x x x
3 x
4 x x
2 x x
-
x x
256/32 Kbytes x x
x x x x x x x
3 x
4 x x
2 x x x x x x x x
1
Package 80, 112-pin
LQFP
80, 112-pin
LQFP
80, 112-pin
LQFP
121 MAPBGA
80, 112-pin
LQFP
121 MAPBGA
112-pin LQFP
121 MAPBGA
The full debug/trace interface is available only on the 112- and 121-pin packages. A reduced debug interface is bonded on the 80-pin package.
x x x x x x x
3 x
4 x x
2
1.2
Block Diagram
The superset device in the MCF52235 family comes in a 112-leaded quad flat package (LQFP) and a 121 pin MAPBGA. Figure 1-1 shows a top-level block diagram of the MCF52235 .
1-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Overview
EPHY_TX
EPHY_RX
EzPD
EzPQ
EPHY
Fast
Ethernet
Controller
(FEC)
4 CH DMA
To/From PADI
JTAG_EN MUX
EzPort
Arbiter
EzPCK
EzPCS
Interrupt
Controller 1
Interrupt
Controller 2
UART
0
DTIM
0
UART
1
UART
2
DTIM
1
DTIM
2
I
2
C
DTIM
3
QSPI
RTC
V2 ColdFire CPU
IFP OEP CAU EMAC
AN[7:0]
JTAG
TAP
ADC
V
RH
V
RL
32 Kbytes
SRAM
(4Kx16)x4
256 Kbytes
Flash
(32Kx16)x4
PORTS
(GPIO)
PMM
CIM
Edge
Port 1
PLL
CLKGEN
Edge
Port 2
EXTAL XTAL CLKOUT
To/From Interrupt Controller
RNGA
FlexCAN
Figure 1-1. MCF52235 Block Diagram
PIT0
PIT1
GPT
PWM
ICOC n
QSPI_DIN,
QSPI_DOUT
QSPI_SCK,
QSPI_PCS n
I
2
C_SDA
I
2
C_SCL
U n TXD
U n RXD
U n RTS
U n CTS
DTIN n /DTOUT n
CANRX
CANTX
PWM n
RSTIN
RSTOUT
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
1-3
Overview
1.3
Part Numbers and Packaging
Table 1-2. Part Number Summary
Part Number
MCF52230
MCF52231
MCF52233
MDCF52234
MCF52235
Flash / SRAM Key Features
128 Kbytes / 32 Kbytes 3 UARTs, I
2
C, QSPI, A/D, FEC EPHY, DMA,
16-/32-bit/PWM Timers
128 Kbytes / 32 Kbytes 3 UARTs, I
2
C, QSPI, A/D, FEC EPHY, DMA,
16-/32-bit/PWM Timers, CAN
256 Kbytes / 32 Kbytes 3 UARTs, I
2
C, QSPI, A/D, FEC EPHY, DMA,
16-/32-bit/PWM Timers
256 Kbytes / 32 Kbytes 3 UARTs, I
2
C, QSPI, A/D, FEC, EPHY, DMA,
16-/32-bit/PWM Timers, CAN
256 Kbytes / 32 Kbytes 3 UARTs, I
2
C, QSPI, A/D, Crypto, FEC,
EPHY, DMA, 16-/32-bit/PWM Timers, CAN
Package
80-pin TQFP
112-pin LQFP
80-pin TQFP
112-pin LQFP
80-pin TQFP
112-pin LQFP
80-pin TQFP
112-pin LQFP
121 MAPBGA
112-pin LQFP
121 MAPBGA
Speed
60 MHz
60 MHz
60 MHz
60 MHz
60 MHz
summarizes the features of the MCF52235 product family. Several speed/package options are available to match cost- or performance-sensitive applications.
1.4
Features
The MCF52235 family includes the following features:
• Version 2 ColdFire variable-length RISC processor core
— Static operation
— 32-bit address and data paths on-chip
— Up to 60 MHz processor core frequency
— Sixteen general-purpose, 32-bit data and address registers
— Implements ColdFire ISA_A with extensions to support the user stack pointer register and four new instructions for improved bit processing (ISA_A+)
— Enhanced Multiply-Accumulate (EMAC) unit with 32-bit accumulator to support
16 × 16 → 32 or 32 × 32 → 32 operations
— Cryptography Acceleration Unit (CAU)
– Tightly-coupled coprocessor to accelerate software-based encryption and message digest functions
– FIPS-140 compliant random number generator
— Support for DES, 3DES, AES, MD5, and SHA-1 algorithms
— Illegal instruction decode that allows for 68K emulation support
• System debug support
— Real time trace for determining dynamic execution path
— Background debug mode (BDM) for in-circuit debugging (DEBUG_B+)
— Real time debug support, with six hardware breakpoints (4 PC, 1 address and 1 data) that can be configured into a 1- or 2-level trigger
1-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Overview
• On-chip memories
— 32-Kbyte dual-ported SRAM on CPU internal bus, supporting core and DMA access with standby power supply support
— 256 Kbytes of interleaved Flash memory supporting 2-1-1-1 accesses
• Power management
— Fully static operation with processor sleep and whole chip stop modes
— Very rapid response to interrupts from the low-power sleep mode (wake-up feature)
— Clock enable/disable for each peripheral when not used
• Fast Ethernet Controller (FEC)
— 10/100 BaseT/TX capability, half duplex or full duplex
— On-chip transmit and receive FIFOs
— Built-in dedicated DMA controller
— Memory-based flexible descriptor rings
• On-chip Ethernet Transceiver (EPHY)
— Digital adaptive equalization
— Supports auto-negotiation
— Baseline wander correction
— Full-/Half-duplex support in all modes
— Loopback modes
— Supports MDIO preamble suppression
— Jumbo packet
• FlexCAN 2.0B module
— Based on and includes all existing features of the Freescale TouCAN module
— Full implementation of the CAN protocol specification version 2.0B
– Standard Data and Remote Frames (up to 109 bits long)
– Extended Data and Remote Frames (up to 127 bits long)
– 0–8 bytes data length
– Programmable bit rate up to 1 Mbit/sec
— Flexible Message Buffers (MBs), totalling up to 16 message buffers of 0–8 byte data length each, configurable as Rx or Tx, all supporting standard and extended messages
— Unused Message Buffer space can be used as general purpose RAM space
— Listen only mode capability
— Content-related addressing
— No read/write semaphores required
— Three programmable mask registers: global for MBs 0-13, special for MB14, and special for
MB15
— Programmable transmit-first scheme: lowest ID or lowest buffer number
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
1-5
Overview
— “Time stamp” based on 16-bit free-running timer
— Global network time, synchronized by a specific message
— Maskable interrupts
• Three universal asynchronous/synchronous receiver transmitters (UARTs)
— 16-bit divider for clock generation
— Interrupt control logic with maskable interrupts
— DMA support
— Data formats can be 5, 6, 7 or 8 bits with even, odd or no parity
— Up to 2 stop bits in 1/16 increments
— Error-detection capabilities
— Modem support includes request-to-send (RTS) and clear-to-send (CTS) lines for two UARTs
— Transmit and receive FIFO buffers
• I 2 C module
— Interchip bus interface for EEPROMs, LCD controllers, A/D converters, and keypads
— Fully compatible with industry-standard I 2 C bus
— Master and slave modes support multiple masters
— Automatic interrupt generation with programmable level
• Queued serial peripheral interface (QSPI)
— Full-duplex, three-wire synchronous transfers
— Up to four chip selects available
— Master mode operation only
— Programmable bit rates up to half the CPU clock frequency
— Up to 16 pre-programmed transfers
• Fast analog-to-digital converter (ADC)
— Eight analog input channels
— 12-bit resolution
— Minimum 1.125 µ s conversion time
— Simultaneous sampling of two channels for motor control applications
— Single-scan or continuous operation
— Optional interrupts on conversion complete, zero crossing (sign change), or under/over low/high limit
— Unused analog channels can be used as digital I/O
• Four 32-bit DMA timers
— 16.7-ns resolution at 60 MHz
— Programmable sources for clock input, including an external clock option
— Programmable prescaler
— Input capture capability with programmable trigger edge on input pin
1-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Overview
— Output compare with programmable mode for the output pin
— Free run and restart modes
— Maskable interrupts on input capture or output compare
— DMA trigger capability on input capture or output compare
• Four-channel general purpose timer
— 16-bit architecture
— Programmable prescaler
— Output pulse widths variable from microseconds to seconds
— Single 16-bit input pulse accumulator
— Toggle-on-overflow feature for pulse-width modulator (PWM) generation
— One dual-mode pulse accumulation channel
• Pulse-width modulation timer
— Operates as eight channels with 8-bit resolution or four channels with 16-bit resolution
— Programmable period and duty cycle
— Programmable enable/disable for each channel
— Software selectable polarity for each channel
— Period and duty cycle are double buffered. Change takes effect when the end of the current period is reached (PWM counter reaches zero) or when the channel is disabled.
— Programmable center or left aligned outputs on individual channels
— Four clock sources (A, B, SA, and SB) provide for a wide range of frequencies
— Emergency shutdown
• Real-Time Clock (RTC)
— Maintains system time-of-day clock
— Provides stopwatch and alarm interrupt functions
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
1-7
Overview
• Two periodic interrupt timers (PITs)
— 16-bit counter
— Selectable as free running or count down
• Software watchdog timer
— 32-bit counter
— Low power mode support
• Clock Generation Features
— 25 MHz crystal input
— On-chip PLL can generate core frequencies up to maximum 60MHz operating frequency
— Provides clock for integrated EPHY
• Dual Interrupt Controllers (INTC0/INTC1)
— Support for multiple interrupt sources organized as follows:
– Fully-programmable interrupt sources for each peripheral
– 7 fixed-level interrupt sources
– Seven external interrupt signals
— Unique vector number for each interrupt source
— Ability to mask any individual interrupt source or all interrupt sources (global mask-all)
— Support for hardware and software interrupt acknowledge (IACK) cycles
— Combinatorial path to provide wake-up from low power modes
• DMA controller
— Four fully programmable channels
— Dual-address transfer support with 8-, 16-, and 32-bit data capability, along with support for
16-byte (4 x 32-bit) burst transfers
— Source/destination address pointers that can increment or remain constant
— 24-bit byte transfer counter per channel
— Auto-alignment transfers supported for efficient block movement
— Bursting and cycle steal support
— Software-programmable DMA requesters for the UARTs (3) and 32-bit timers (4)
• Reset
— Separate reset in and reset out signals
— Seven sources of reset:
– Power-on reset (POR)
– External
– Software
– Watchdog
– Loss of clock
– Loss of lock
1-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Overview
– Low-voltage detection (LVD)
— Status flag indication of source of last reset
• Chip integration module (CIM)
— System configuration during reset
— Selects one of three clock modes
— Configures output pad drive strength
— Unique part identification number and part revision number
• General purpose I/O interface
— Up to 56 bits of general purpose I/O
— Bit manipulation supported via set/clear functions
— Programmable drive strengths
— Unused peripheral pins may be used as extra GPIO
• JTAG support for system level board testing
1.4.1
V2 Core Overview
The version 2 ColdFire processor core is comprised of two separate pipelines that are decoupled by an instruction buffer. The two-stage instruction fetch pipeline (IFP) is responsible for instruction-address generation and instruction fetch. The instruction buffer is a first-in-first-out (FIFO) buffer that holds prefetched instructions awaiting execution in the operand execution pipeline (OEP). The OEP includes two pipeline stages. The first stage decodes instructions and selects operands (DSOC); the second stage
(AGEX) performs instruction execution and calculates operand effective addresses, if needed.
The V2 core implements the ColdFire instruction set architecture revision A+ with added support for a separate user stack pointer register and four new instructions to assist in bit processing. Additionally, the
MCF52235 core includes the enhanced multiply-accumulate (EMAC) unit for improved signal processing capabilities. The MAC implements a three-stage arithmetic pipeline, optimized for 16 x 16 bit operations, with support for one 32-bit accumulator. Supported operands include 16- and 32-bit signed and unsigned integers, signed fractional operands, and a complete set of instructions to process these data types. The EMAC provides support for execution of DSP operations within the context of a single processor at a minimal hardware cost.
1.4.2
Integrated Debug Module
The ColdFire processor core debug interface is provided to support system debugging in conjunction with low-cost debug and emulator development tools. Through a standard debug interface, users can access debug information and real-time tracing capability is provided on 112-and 121-lead packages. This allows the processor and system to be debugged at full speed without the need for costly in-circuit emulators.
The on-chip breakpoint resources include a total of nine programmable 32-bit registers: an address and an address mask register, a data and a data mask register, four PC registers, and one PC mask register. These registers can be accessed through the dedicated debug serial communication channel or from the processor’s supervisor mode programming model. The breakpoint registers can be configured to generate triggers by combining the address, data, and PC conditions in a variety of single- or dual-level definitions.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
1-9
Overview
The trigger event can be programmed to generate a processor halt or initiate a debug interrupt exception.
The MCF52235 implements revision B+ of the coldfire Debug Architecture.
The MCF52235 ’s interrupt servicing options during emulator mode allow real-time critical interrupt service routines to be serviced while processing a debug interrupt event, thereby ensuring that the system continues to operate even during debugging.
To support program trace, the V2 debug module provides processor status (PST[3:0]) and debug data
(DDATA[3:0]) ports. These buses and the PSTCLK output provide execution status, captured operand data, and branch target addresses defining processor activity at the CPU’s clock rate. The MCF52235 includes a new debug signal, ALLPST. This signal is the logical ‘AND’ of the processor status (PST[3:0]) signals and is useful for detecting when the processor is in a halted state (PST[3:0] = 1111).
The full debug/trace interface is available only on the 112 and 121-pin packages. However, every product features the dedicated debug serial communication channel (DSI, DSO, DSCLK) and the ALLPST signal.
1.4.3
JTAG
The MCF52235 supports circuit board test strategies based on the Test Technology Committee of IEEE and the Joint Test Action Group (JTAG). The test logic includes a test access port (TAP) consisting of a
16-state controller, an instruction register, and three test registers (a 1-bit bypass register, a -bit boundary-scan register, and a 32-bit ID register). The boundary scan register links the device’s pins into one shift register. Test logic, implemented using static logic design, is independent of the device system logic.
The MCF52235 implementation can do the following:
• Perform boundary-scan operations to test circuit board electrical continuity
• Sample MCF52235 system pins during operation and transparently shift out the result in the boundary scan register
• Bypass the MCF52235 for a given circuit board test by effectively reducing the boundary-scan register to a single bit
• Disable the output drive to pins during circuit-board testing
• Drive output pins to stable levels
1.4.4
On-Chip Memories
1.4.4.1
SRAM
The dual-ported SRAM module provides a general-purpose 32-Kbyte memory block that the ColdFire core can access in a single cycle. The location of the memory block can be set to any 32-Kbyte boundary within the 4-Gbyte address space. This memory is ideal for storing critical code or data structures and for use as the system stack. Because the SRAM module is physically connected to the processor's high-speed local bus, it can quickly service core-initiated accesses or memory-referencing commands from the debug module.
1-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Overview
The SRAM module is also accessible by the DMA. The dual-ported nature of the SRAM makes it ideal for implementing applications with double-buffer schemes, where the processor and a DMA device operate in alternate regions of the SRAM to maximize system performance.
1.4.4.2
Flash
The ColdFire Flash module (CFM) is a non-volatile memory (NVM) module that connects to the processor’s high-speed local bus. The CFM is constructed with four banks of 32K x 16-bit Flash arrays to generate 256 Kbytes of 32-bit Flash memory. These arrays serve as electrically erasable and programmable, non-volatile program and data memory. The Flash memory is ideal for program and data storage for single-chip applications, allowing for field reprogramming without requiring an external high voltage source. The CFM interfaces to the ColdFire core through an optimized read-only memory controller which supports interleaved accesses from the 2-cycle Flash arrays. A backdoor mapping of the
Flash memory is used for all program, erase, and verify operations, as well as providing a read datapath for the DMA. Flash memory may also be programmed via the EzPort, which is a serial Flash programming interface that allows the Flash to be read, erased and programmed by an external controller in a format compatible with most SPI bus Flash memory chips. This allows easy device programming via Automated
Test Equipment or bulk programming tools.
1.4.5
Cryptography Acceleration Unit
The MCF52235 device incorporates two hardware accelerators for cryptographic functions. First, the
CAU is a coprocessor tightly-coupled to the V2 ColdFire core that implements a set of specialized operations to increase the throughput of software-based encryption and message digest functions, specifically the DES, 3DES, AES, MD5 and SHA-1 algorithms. Second, a random number generator provides FIPS-140 compliant 32-bit values to security processing routines. Both modules supply critical acceleration to software-based cryptographic algorithms at a minimal hardware cost.
1.4.6
Power Management
The MCF52235 incorporates several low power modes of operation which are entered under program control and exited by several external trigger events. An integrated power-on reset (POR) circuit monitors the input supply and forces an MCU reset as the supply voltage rises. The low voltage detector (LVD) monitors the supply voltage and is configurable to force a reset or interrupt condition if it falls below the
LVD trip point.
1.4.7
FlexCAN
The FlexCAN module is a communication controller implementing version 2.0 of the CAN protocol parts
A and B. The CAN protocol can be used as an industrial control serial data bus, meeting the specific requirements of reliable operation in a harsh EMI environment with high bandwidth. This instantiation of
FlexCAN has 16 message buffers.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
1-11
Overview
1.4.8
UARTs
The MCF52235 has three full-duplex UARTs that function independently. The three UARTs can be clocked by the system bus clock, eliminating the need for an external clock source. On smaller packages, the third UART is multiplexed with other digital I/O functions.
1.4.9
I
2
C Bus
The I 2 C bus is a two-wire, bidirectional serial bus that provides a simple, efficient method of data exchange and minimizes the interconnection between devices. This bus is suitable for applications requiring occasional communications over a short distance between many devices on a circuit board.
1.4.10
QSPI
The queued serial peripheral interface (QSPI) provides a synchronous serial peripheral interface with queued transfer capability. It allows up to 16 transfers to be queued at once, minimizing the need for CPU intervention between transfers.
1.4.11
Fast ADC
The Fast ADC consists of an eight-channel input select multiplexer and two independent sample and hold
(S/H) circuits feeding separate 12-bit ADCs. The two separate converters store their results in accessible buffers for further processing.
The ADC can be configured to perform a single scan and halt, perform a scan whenever triggered, or perform a programmed scan sequence repeatedly until manually stopped.
The ADC can be configured for either sequential or simultaneous conversion. When configured for sequential conversions, up to eight channels can be sampled and stored in any order specified by the channel list register. Both ADCs may be required during a scan, depending on the inputs to be sampled.
During a simultaneous conversion, both S/H circuits are used to capture two different channels at the same time. This configuration requires that a single channel may not be sampled by both S/H circuits simultaneously.
Optional interrupts can be generated at the end of the scan sequence if a channel is out of range (measures below the low threshold limit or above the high threshold limit set in the limit registers) or at several different zero crossing conditions.
1.4.12
DMA Timers (DTIM0–DTIM3)
There are four independent, DMA transfer capable 32-bit timers (DTIM0, DTIM1, DTIM2, and DTIM3) on the each device. Each module incorporates a 32-bit timer with a separate register set for configuration and control. The timers can be configured to operate from the system clock or from an external clock source using one of the DTINx signals. If the system clock is selected, it can be divided by 16 or 1. The input clock is further divided by a user-programmable 8-bit prescaler which clocks the actual timer counter register (TCRn). Each of these timers can be configured for input capture or reference (output) compare mode. Timer events may optionally cause interrupt requests or DMA transfers.
1-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Overview
1.4.13
General Purpose Timer (GPT)
The general purpose timer (GPT) is a 4-channel timer module consisting of a 16-bit programmable counter driven by a 7-stage programmable prescaler. Each of the four channels can be configured for input capture or output compare. Additionally, one of the channels, channel 3, can be configured as a pulse accumulator.
A timer overflow function allows software to extend the timing capability of the system beyond the 16-bit range of the counter. The input capture and output compare functions allow simultaneous input waveform measurements and output waveform generation. The input capture function can capture the time of a selected transition edge. The output compare function can generate output waveforms and timer software delays. The 16-bit pulse accumulator can operate as a simple event counter or a gated time accumulator.
1.4.14
Periodic Interrupt Timers (PIT0 and PIT1)
The two periodic interrupt timers (PIT0 and PIT1) are 16-bit timers that provide interrupts at regular intervals with minimal processor intervention. Each timer can either count down from the value written in its PIT modulus register, or it can be a free-running down-counter.
1.4.15
Pulse Width Modulation (PWM) Timers
The MCF52235 has an 8-channel, 8-bit PWM timer. Each channel has a programmable period and duty cycle as well as a dedicated counter. Each of the modulators can create independent continuous waveforms with software-selectable duty rates from 0% to 100%. The PWM outputs have programmable polarity, and can be programmed as left aligned outputs or center aligned outputs. For higher period and duty cycle resolution, each pair of adjacent channels ([7:6], [5:4], [3:2], and [1:0]) can be concatenated to form a single 16-bit channel. The module can thus be configured to support 8/0, 6/1, 4/2, 2/3, or 0/4 8-/16-bit channels.
1.4.16
Software Watchdog Timer
The watchdog timer is a 32-bit timer that facilitates recovery from runaway code. The watchdog counter is a free-running down-counter that generates a reset on underflow. To prevent a reset, software must periodically restart the countdown.
1.4.17
Phase Locked Loop (PLL)
The clock module contains a crystal oscillator, 8 MHz on-chip relaxation oscillator (OCO), phase-locked loop (PLL), reduced frequency divider (RFD), low-power divider status/control registers, and control logic. In order to improve noise immunity, the PLL, crystal oscillator, and relaxation oscillator have their own power supply inputs: VDDPLL and VSSPLL. All other circuits are powered by the normal supply pins, VDD and VSS.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
1-13
Overview
1.4.18
Interrupt Controller (INTC0/INTC1)
There are two interrupt controllers on the MCF52235. These interrupt controllers are organized as seven levels with up to nine interrupt sources per level. Each interrupt source has a unique interrupt vector, and provide each peripheral with all necessary interrupts. Each internal interrupt has a programmable level
[1-7] and priority within the level. The seven external interrupts have fixed levels/priorities.
1.4.19
DMA Controller
The direct memory access (DMA) controller provides an efficient way to move blocks of data with minimal processor intervention. It has four channels that allow byte, word, longword, or 16-byte burst line transfers. These transfers are triggered by software explicitly setting a DCRn[START] bit or by the occurrence of certain UART or DMA timer events.
1.4.20
Reset
The reset controller determines the source of reset, asserts the appropriate reset signals to the system, and keeps track of what caused the last reset. There are seven sources of reset:
• External reset input
• Power-on reset (POR)
• Watchdog timer
• Phase locked-loop (PLL) loss of lock
• PLL loss of clock
• Software
• Low-voltage detector (LVD)
Control of the LVD and its associated reset and interrupt are handled by the reset controller. Other registers provide status flags indicating the last source of reset and a control bit for software assertion of the RSTO pin.
1.4.21
GPIO
Nearly all pins on the MCF52235 have general purpose I/O capability in addition to their primary functions, and are grouped into 8-bit ports. Some ports do not utilize all 8 bits. Each port has registers that configure, monitor, and control the port pins.
1-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Signal Descriptions
Chapter 2
Signal Descriptions
2.1
Introduction
This chapter describes signals implemented on this device and includes an alphabetical listing of signals that characterizes each signal as an input or output, defines its state at reset, and identifies whether a pull-up resistor should be used.
NOTE
The terms ‘assertion’ and ‘negation’ are used to avoid confusion when dealing with a mixture of active-low and active-high signals. The term
‘asserted’ indicates that a signal is active, independent of the voltage level.
The term ‘negated’ indicates that a signal is inactive.
Active-low signals, such as SRAS and TA, are indicated with an overbar.
2.2
Overview
shows the block diagram of the device with the signal interface.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
2-1
Signal Descriptions
EPHY_TX
EPHY_RX
EzPD
EzPQ
EPHY
Fast
Ethernet
Controller
(FEC)
4 CH DMA
To/From PADI
JTAG_EN MUX
EzPort
Arbiter
EzPCK
EzPCS
Interrupt
Controller 1
Interrupt
Controller 2
UART
0
DTIM
0
UART
1
UART
2
DTIM
1
DTIM
2
I
2
C
DTIM
3
QSPI
RTC
V2 ColdFire CPU
IFP OEP CAU EMAC
AN[7:0]
JTAG
TAP
ADC
V
RH
V
RL
32 Kbytes
SRAM
(4Kx16)x4
256 Kbytes
Flash
(32Kx16)x4
PORTS
(GPIO)
PMM
CIM
ICOC n
QSPI_DIN,
QSPI_DOUT
QSPI_SCK,
QSPI_PCS n
I
2
C_SDA
I
2
C_SCL
U n TXD
U n RXD
U n RTS
U n CTS
DTIN n /DTOUT n
CANRX
CANTX
PWM n
RSTIN
RSTOUT
Edge
Port 1
PLL
CLKGEN
Edge
Port 2
EXTAL XTAL CLKOUT
To/From Interrupt Controller
RNGA
FlexCAN
PIT0
PIT1
GPT
PWM
Figure 2-1. Block Diagram with Signal Interfaces
shows the pin functions by primary and alternate purpose, and illustrates which packages contain each pin.
2-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Pin
Group
Primary
Function
ADC
Clock
Generation
Debug
Data
SYNCA
SYNCB
VDDA
VSSA
VRH
VRL
EXTAL
XTAL
AN3
AN2
AN1
AN0
AN7
AN6
AN5
AN4
VDDPLL
VSSPLL
ALLPST
DDATA[3:0]
PST[3:0]
—
—
—
—
—
—
—
—
CANTX
3
CANRX
3
—
—
—
—
—
—
—
—
—
—
Secondary
Function
—
Table 2-1. Pin Functions by Primary and Alternate Purpose
Tertiary
Function
FEC_MDIO
FEC_MDC
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
Quaternary
Function
Drive
Strength /
Control
1
Wired OR
Control
Pull-up /
Pull-down
2
Pin on121
MAPBGA
Pin on 112
LQFP
Pin on 80
LQFP
PDSR[39]
PDSR[39]
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
High
High
Low
Low
Low
Low
Low
Low
Low
Low
—
—
—
—
PAS[3]
PAS[2]
—
—
PAN[7]
PAN[6]
PAN[5]
PAN[4]
PAN[3]
PAN[2]
PAN[1]
PAN[0]
—
—
—
PDD[7:4]
—
—
—
—
—
—
—
—
—
—
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
PDD[3:0] High — — —
95
96
48
49
28
27
93
94
45
47
7
12,13,
16,17
80,79,
78,77
89
90
91
92
88
87
86
85
70
71
36
37
20
19
69
72
33
35
7
—
65
66
67
68
64
63
62
61
—
Notes
Pin
Group
Primary
Function
Ethernet
LEDs
Ethernet
PHY
I
2
C
ACTLED
COLLED
DUPLED
LNKLED
SPDLED
RXLED
TXLED
PHY_RBIAS
PHY_RXN
PHY_RXP
PHY_TXN
PHY_TXP
PHY_VDDA
PHY_VDDRX
PHY_VDDTX
PHY_VSSA
PHY_VSSRX
PHY_VSSTX
SCL
SDA
—
—
—
—
—
—
—
—
—
—
CANTX
3
CANRX
3
—
—
—
—
—
—
—
—
Secondary
Function
—
—
—
—
—
—
—
—
—
—
TXD2
RXD2
—
—
—
—
—
—
—
—
Tertiary
Function
Quaternary
Function
Drive
Strength /
Control
1
Wired OR
Control
Pull-up /
Pull-down
2
Pin on121
MAPBGA
Pin on 112
LQFP
Pin on 80
LQFP
—
—
—
—
—
—
—
—
PLD[0]
PLD[4]
PLD[3]
PLD[1]
PLD[2]
PLD[5]
PLD[6]
—
—
—
PAS[0]
PAS[1]
PDSR[32] PWOR[8]
PDSR[36] PWOR[12]
PDSR[35] PWOR[11]
PDSR[33] PWOR[9]
PDSR[34] PWOR[10]
PDSR[37] PWOR[13]
PDSR[38] PWOR[14]
— —
—
—
—
—
PDSR[0]
PDSR[0]
N/A
N/A
—
—
N/A
N/A
N/A
N/A
—
—
—
—
—
—
—
—
—
—
— pull-up pull-up
4
4
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
68
75
69
67
74
73
71
70
76
72
111
112
81
52
51
66
84
58
59
83
48
55
49
47
54
53
51
50
56
52
79
80
57
—
—
46
60
42
43
59
Notes
Pin
Group
Primary
Function
Interrupts
JTAG/BDM
Mode
Selection
IRQ7
IRQ6
IRQ5
IRQ4
IRQ3
IRQ2
IRQ1
JTAG_EN
TCLK/
PSTCLK
TDI/DSI
TDO/DSO
TMS
/BKPT
TRST
/DSCLK
RCON/
EZPCS
IRQ15
IRQ14
IRQ13
IRQ12
IRQ11
IRQ10
IRQ9
IRQ8
Secondary
Function
—
—
—
—
—
—
—
—
—
—
—
—
—
—
SYNCA
—
CLKOUT
Tertiary
Function
—
—
—
—
—
—
—
—
—
FEC_RXER
FEC_RXD[1]
—
FEC_RXD[2]
FEC_RXD[3]
PWM1
—
—
Quaternary
Function
Drive
Strength /
Control
1
Wired OR
Control
Pull-up /
Pull-down
2
Pin on121
MAPBGA
Pin on 112
LQFP
Pin on 80
LQFP
PGP[7]
PGP[6]
PGP[5]
PGP[4]
PGP[3]
PGP[2]
PGP[1]
PGP[0]
PNQ[7]
PNQ[6]
PNQ[5]
PNQ[4]
PNQ[3]
PNQ[2]
PNQ[1]
—
—
PSDR[47]
PSDR[46]
PSDR[45]
PSDR[44]
PSDR[43]
PSDR[42]
PSDR[41]
PSDR[40]
Low
Low
Low
Low
Low
Low
High
N/A
High
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
N/A
— pull-up
4 pull-up
4 pull-up
4 pull-up
4 pull-up
4 pull-up
4 pull-up
4 pull-up
4 pull-up
4 pull-up
4 pull-up
4 pull-up
4 pull-up
4 pull-up
4 pull-up
4 pull-down pull-up
5
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
106
105
98
97
57
29
11
10
56
19
20
41
53
54
55
18
1
—
—
—
—
—
—
—
—
40
—
—
29
—
—
39
12
1
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
N/A
High
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A pull-up
—
5 pull-up
5 pull-up pull-up
5
—
—
—
—
—
4
5
2
6
3
4
5
2
6
3
Notes
Pin
Group
Primary
Function
Secondary
Function
PWM
QSPI
Reset
7
Test
Timers,
16-bit
Timers,
32-bit
PWM7
PWM5
PWM3
PWM1
QSPI_DIN/
EZPD
QSPI_DOUT/
EZPQ
QSPI_SCK/
EZPCK
QSPI_CS3
QSPI_CS2
QSPI_CS1
QSPI_CS0
RSTI
RSTO
TEST
GPT3
—
—
—
—
CANRX
3
CANTX
SCL
3
SYNCA
—
—
SDA
—
—
—
FEC_TXD[3]
GPT2
GPT1
GPT0
TIN3
TIN2
TIN1
TIN0
FEC_TXD[2]
FEC_TXD[1]
FEC_TXER
TOUT3
TOUT2
TOUT1
TOUT0
Tertiary
Function
—
—
—
—
RXD1
Quaternary
Function
Drive
Strength /
Control
1
Wired OR
Control
Pull-up /
Pull-down
2
Pin on121
MAPBGA
Pin on 112
LQFP
Pin on 80
LQFP
PTD[3]
PTD[2]
PTD[1]
PTD[0]
PQS[1]
PDSR[31]
PDSR[30]
PDSR[29]
PDSR[28]
—
—
—
—
PDSR[2] PWOR[4]
—
—
—
—
—
—
—
—
—
—
104
103
100
99
34
—
—
—
—
25
TXD1
RTS1
SYNCB
—
—
CTS1
—
—
—
PWM7
PWM5
PWM3
PWM1
PWM6
PWM4
PWM2
PWM0
PQS[0]
PQS[2]
PQS[6]
PQS[5]
PQS[4]
PQS[3]
—
—
—
PTA[3]
PTA[2]
PTA[1]
PTA[0]
PTC[3]
PTC[2]
PTC[1]
PTC[0]
PDSR[1]
PDSR[3]
PDSR[7]
PDSR[6]
PWOR[5]
PWOR[6]
PDSR[5] —
PDSR[4] PWOR[7]
N/A high
N/A
PDSR[23]
PDSR[22]
PDSR[21]
PDSR[20]
PDSR[19]
PDSR[18]
PDSR[17]
PDSR[16]
—
—
N/A
—
N/A
—
—
—
—
—
—
—
—
— pull-up
—
—
6
— pull-up
6 pull-up
7
— pull-down pull-up
8 pull-up
8 pull-up
8 pull-up
8
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
35
36
40
39
38
37
44
46
50
107
108
109
110
22
21
9
8
26
27
13
9
8
76
77
78
14
32
34
38
75
—
—
—
58
Notes
Pin
Group
Primary
Function
Secondary
Function
Tertiary
Function
Quaternary
Function
Drive
Strength /
Control
1
Wired OR
Control
Pull-up /
Pull-down
2
Pin on121
MAPBGA
Pin on 112
LQFP
Pin on 80
LQFP
Notes
UART 0
UART 1
CTS0
RTS0
RXD0
TXD0
CTS1
RTS1
RXD1
TXD1
CANRX
CANTX
—
—
SYNCA
SYNCB
—
—
3
3
FEC_RXCLK
FEC_RXDV
FEC_RXD[0]
FEC_CRS
RXD2
TXD2
FEC_TXD[0]
FEC_COL
PUA[3]
PUA[2]
PUA[1]
PUA[0]
PUB[3]
PUB[2]
PUB[1]
PUB[0]
PDSR[11]
PDSR[10]
PDSR[9]
PDSR[8]
PDSR[15]
PDSR[14]
PDSR[13]
PDSR[12]
—
—
PWOR[0]
PWOR[1]
—
—
PWOR[2]
PWOR[3]
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
26
25
30
31
24
23
32
33
18
17
21
22
16
15
23
24
UART 2 CTS2
RTS2
—
—
—
—
PUC[3]
PUC[2]
PDSR[27]
PDSR[26]
—
—
—
—
—
—
61
60
—
—
FlexCAN
RXD2
TXD2
SYNCA
SYNCB
—
—
CANTX
9
CANRX
3
—
—
—
FEC_MDIO
FEC_MDC
PUC[1]
PUC[30]
PAS[3]
PAS[2]
PDSR[25]
PDSR[24]
PDSR[39]
PDSR[39]
—
—
—
—
—
—
—
—
—
—
—
—
62
63
—
—
—
—
—
—
See Note
See Note
3
3
VDD
10
VDD — — N/A N/A — 14,43,65,
82,102
10,31,45,
58,74
VSS VSS — — — N/A N/A — 15,42,
64,101
11,30,
44,73
2
3
4
5
6
7
8
9
1
10
mode.
All signals have a pull-up in GPIO mode.
The multiplexed CANTX and CANRX signals do not have dedicated pins, but are available as muxed replacements for other signals.
For primary and GPIO functions only.
Only when JTAG mode is enabled.
For secondary and GPIO functions only.
RSTI has an internal pull-up resistor, however the use of an external resistor is very strongly recommended
For GPIO function. Primary Function has pull-up control within the GPT module
The multiplexed CANTX and CANRX signals do not have dedicated pins, but are available as muxed replacements for other signals.
This list for power and ground does not include those dedicated power/ground pins included elsewhere, e.g. in the ethernet PHY.
Signal Descriptions
2.3
Reset Signals
describes signals that are used to either reset the chip or as a reset indication.
Table 2-2. Reset Signals
Signal Name
Reset In
Reset Out
Abbreviation
RSTI
RSTO
Function
Primary reset input to the device. Asserting RSTI immediately resets the CPU and peripherals.
Driven low for 512 CPU clocks after the reset source has deasserted and PLL locked.
I/O
I
O
2.4
PLL and Clock Signals
describes signals that are used to support the on-chip clock generation circuitry.
Table 2-3. PLL and Clock Signals
Signal Name
External Clock In
Crystal
Clock Out
Abbreviation Function
EXTAL
XTAL
Crystal oscillator or external clock input.
Crystal oscillator output.
CLKOUT This output signal reflects the internal system clock.
I/O
I
O
O
2.5
Mode Selection
describes signals used in mode selection, Table 2-5 describes particular clocking modes.
Table 2-4. Mode Selection Signals
Signal Name
Reset Configuration
Test
Abbreviation
RCON
TEST
Function
The serial Flash programming mode is entered by asserting the
RCON pin (with the TEST pin negated) as the chip comes out of reset. During this mode, the EzPort has access to the Flash memory which can be programmed from an external device.
Reserved for factory testing only and in normal modes of operation should be connected to VSS to prevent unintentional activation of test functions.
I/O
I
2.6
External Interrupt Signals
describes the external interrupt signals.
Table 2-5. External Interrupt Signals
Signal Name
External Interrupts
Abbreviation
IRQ[15:1] External interrupt sources.
Function I/O
I
2-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Signal Descriptions
2.7
Queued Serial Peripheral Interface (QSPI)
describes the QSPI signals.
Table 2-6. Queued Serial Peripheral Interface (QSPI) Signals
Signal Name Abbreviation Function
QSPI Synchronous
Serial Output
QSPI Synchronous
Serial Data Input
QSPI Serial Clock
QSPI_DOUT Provides the serial data from the QSPI and can be programmed to be driven on the rising or falling edge of QSPI_CLK.
QSPI_DIN Provides the serial data to the QSPI and can be programmed to be sampled on the rising or falling edge of QSPI_CLK.
QSPI_CLK Provides the serial clock from the QSPI. The polarity and phase of
QSPI_CLK are programmable.
Synchronous Peripheral
Chip Selects
QSPI_CS[3:0] QSPI peripheral chip selects that can be programmed to be active high or low.
I/O
O
I
O
O
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
2-9
Signal Descriptions
2.8
Fast Ethernet Controller PHY Signals
describes the Fast Ethernet Controller (FEC) Signals.
Table 7. Fast Ethernet Controller (FEC) Signals
Signal Name
Twisted Pair Input +
Twisted Pair Input -
Twisted Pair Output +
Twisted Pair Output -
Bias Control Resistor
Activity LED
Link LED
Speed LED
Duplex LED
Collision LED
Transmit LED
Receive LED
Abbreviation Function
RXP
RXN
TXN
Differential Ethernet twisted-pair input pin. This pin is high-impedance out of reset.
Differential Ethernet twisted-pair input pin. This pin is high-impedance out of reset.
Differential Ethernet twisted-pair output pin. This pin is high-impedance out of reset.
TXP
RBIAS
Differential Ethernet twisted-pair output pin. This pin is high-impedance out of reset.
Connect a 12.4 k
Ω
(1.0%) external resistor, RBIAS, between the
PHY_RBIAS pin and analog ground.
Place this resistor as near to the chip pin as possible. Stray capacitance must be kept to less than 10 pF
(>50 pF will cause instability). No high-speed signals can be permitted in the region of RBIAS.
ACT_LED Indicates when the EPHY is transmitting or receiving
LINK_LED Indicates when the EPHY has a valid link
SPD_LED Indicates the speed of the EPHY connection
DUPLED Indicates the duplex (full or half) of the EPHY connection
COLLED Indicates if the EPHY detects a collision
TXLED Indicates if the EPHY is transmitting
RXLED Indicates if the EPHY is receiving
I/O
I
I
I
O
O
O
O
O
O
O
O
O
2.9
I
2
C I/O Signals
describes the I
2
C serial interface module signals.
Table 2-8. I
2
C I/O Signals
Signal Name
Serial Clock
Serial Data
Abbreviation
SCL
SDA
Function
Open-drain clock signal for the for the I
2
C interface. Either it is driven by the I
2
C module when the bus is in master mode or it becomes the clock input when the I
2
C is in slave mode.
Open-drain signal that serves as the data input/output for the I
2
C interface.
I/O
I/O
I/O
2-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Signal Descriptions
2.10
UART Module Signals
describes the UART module signals.
Table 2-9. UART Module Signals
Signal Name
Transmit Serial Data Output
Abbreviation
UTXD n
Receive Serial Data Input
Clear-to-Send
Request-to-Send
URXD
UCTS
URTS n n n
Function I/O
Transmitter serial data outputs for the UART modules. The output is held high (mark condition) when the transmitter is disabled, idle, or in the local loopback mode. Data is shifted out, LSB first, on this pin at the falling edge of the serial clock source.
Receiver serial data inputs for the UART modules. Data is received on this pin LSB first. When the UART clock is stopped for power-down mode, any transition on this pin restarts it.
O
I
Indicate to the UART modules that they can begin data transmission.
I
Automatic request-to-send outputs from the UART modules. This signal can also be configured to be asserted and negated as a function of the RxFIFO level.
O
2.11 DMA Timer Signals
describes the signals of the four DMA timer modules.
Table 2-10. DMA Timer Signals
Signal Name
DMA Timer Input
DMA Timer Output
Abbreviation
DTIN
DTOUT
Function
Event input to the DMA timer modules.
Programmable output from the DMA timer modules.
2.12 ADC Signals
Table 2-11 describes the signals of the analog-to-digital converter.
Table 2-11. ADC Signals
Signal Name
Analog Inputs
Analog Reference
Analog Supply
Abbreviation
AN[7:0]
V
RH
V
RL
V
DDA
V
SSA
Function
Inputs to the A-to-D converter.
Reference voltage high and low inputs.
Isolate the ADC circuitry from power supply noise
I/O
I
I
I
—
—
I/O
I
O
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
2-11
Signal Descriptions
2.13 General Purpose Timer Signals
describes the general purpose timer signals.
Table 2-12. GPT Signals
Signal Name
General Purpose Timer
Input/Output
Abbreviation
GPT[3:0]
Function
Inputs to or outputs from the general purpose timer module
2.14 Pulse Width Modulator Signals
describes the PWM signals.
Table 2-13. PWM Signals
Signal Name
PWM Output Channels
Abbreviation
PWM[7:0]
Function
Pulse width modulated output for PWM channels
I/O
I/O
I/O
O
2.15 Debug Support Signals
The signals in
are used as the interface to the on-chip JTAG controller and also to interface to the BDM logic.
Table 2-14. Debug Support Signals
Signal Name
JTAG Enable
Test Reset
Test Clock
Test Mode Select
Test Data Input
Test Data Output
Development Serial
Clock
Breakpoint
Abbreviation Function
JTAG_EN Select between debug module and JTAG signals at reset
TRST This active-low signal is used to initialize the JTAG logic asynchronously.
TCLK
TMS
TDI
Used to synchronize the JTAG logic.
Used to sequence the JTAG state machine. TMS is sampled on the rising edge of TCLK.
Serial input for test instructions and data. TDI is sampled on the rising edge of TCLK.
TDO
DSCLK
BKPT
Serial output for test instructions and data. TDO is three-stateable and is actively driven in the shift-IR and shift-DR controller states. TDO changes on the falling edge of TCLK.
Development Serial Clock. Internally synchronized input. (The logic level on DSCLK is validated if it has the same value on two consecutive rising bus clock edges.) Clocks the serial communication port to the debug module during packet transfers. Maximum frequency is PSTCLK/5. At the synchronized rising edge of DSCLK, the data input on DSI is sampled and DSO changes state.
Breakpoint. Input used to request a manual breakpoint. Assertion of
BKPT puts the processor into a halted state after the current instruction completes. Halt status is reflected on processor status signals () as the value 0xF.
I
I
I/O
I
I
I
I
I
O
2-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Signal Descriptions
Table 2-14. Debug Support Signals (continued)
Signal Name Abbreviation Function
Development Serial
Input
Development Serial
Output
Debug Data
Processor Status Clock
Processor Status
Outputs
All Processor Status
Outputs
DSI
DSO
Development Serial Input. Internally synchronized input that provides data input for the serial communication port to the debug module, once the DSCLK has been seen as high (logic 1).
Development Serial Output. Provides serial output communication for debug module responses. DSO is registered internally. The output is delayed from the validation of DSCLK high.
DDATA[3:0] Debug data. Displays captured processor data and breakpoint status.
The CLKOUT signal can be used by the development system to know when to sample DDATA[3:0].
PSTCLK Processor Status Clock. Delayed version of the processor clock. Its rising edge appears in the center of valid PST and DDATA output.
PSTCLK indicates when the development system should sample PST and DDATA values.
If real-time trace is not used, setting CSR[PCD] keeps PSTCLK, and
PST and DDATA outputs from toggling without disabling triggers.
Non-quiescent operation can be reenabled by clearing CSR[PCD], although the external development systems must resynchronize with the PST and DDATA outputs.
PSTCLK starts clocking only when the first non-zero PST value (0xC,
0xD, or 0xF) occurs during system reset exception processing.
PST[3:0] Indicate core status. Debug mode timing is synchronous with the processor clock; status is unrelated to the current bus transfer. The
CLKOUT signal can be used by the development system to know when to sample PST[3:0].
ALLPST Logical “AND” of PST[3.0]
I/O
I
O
O
O
O
O
2.16 EzPort Signal Descriptions
contains a list of EzPort external signals
Table 2-15. EzPort Signal Descriptions
Signal Name
EzPort Clock
EzPort Chip Select
EzPort Serial Data In
EzPort Serial Data Out
Abbreviation
EZPCK
EZPCS
EZPD
EZPQ
Function
Shift clock for EzPort transfers
Chip select for signaling the start and end of serial transfers
EZPD is sampled on the rising edge of EZPCK
EZPQ transitions on the falling edge of EZPCK
I/O
I
I
I
O
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
2-13
Signal Descriptions
2.17 Power and Ground Pins
for adequate current capability. All power supply pins must have adequate decoupling (bypass capacitance) for high-frequency noise suppression.
Table 2-16. Power and Ground Pins
Signal Name
PLL Analog Supply
Positive Supply
Ground
Abbreviation
VDDPLL,
VSSPLL
VDD
VSS
Function
Dedicated power supply signals to isolate the sensitive PLL analog circuitry from the normal levels of noise present on the digital power supply.
These pins supply positive power to the core logic.
This pin is the negative supply (ground) to the chip.
I/O
I
I
Some of the V
DD
and V
SS
pins on the device are only to be used for noise bypass. Figure 2 shows a typical
connection diagram. Pay particular attention to those pins which show only capacitor connections. Do not connect power supply voltage directly to these pins unless the desire is to send the device into a slow but certain death spiral.
2-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Signal Descriptions
MCF52235
Pin numbering is shown for the 80-lead TQFP
45
44
74
73
10
31
30
71
72
58
11
33
35
69
70
V
DD
PLL
V
SS
PLL
V
DDA
V
RH
V
RL
V
SSA
V
DDR
V
SSX1
V
DDX1
V
DDX2
V
SSX2
V
DD2
V
SS2
V
DD1
V
SS1
0.22µF 1000pF
0.1µF
0.1µF
*
10µH
10µF
10V
Tantalum
0.1µF
0.1µF
0.1µF
0.22µF
3.3V
0.22µF
12.4K
Ω
1%
0.22µF 0.22µF 0.22µF
*
optional
Figure 2. Suggested connection scheme for Power and Ground
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
2-15
Signal Descriptions
2-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 3
ColdFire Core
This section describes the organization of the Version 2 (V2) ColdFire
®
processor core and an overview of the program-visible registers. For detailed information on instructions, see the ISA_A+ definition in the
ColdFire Family Programmer’s Reference Manual .
3.1
Processor Pipelines
is a block diagram showing the processor pipelines of a V2 ColdFire core.
Instruction
Fetch
Pipeline
IAG
IC
Instruction
Address
Generation
Instruction
Fetch Cycle
Address [31:0]
IB
FIFO
Instruction Buffer
Data[31:0]
Operand
Execution
Pipeline
DSOC
Decode & Select,
Operand Fetch
AGEX
Address
Generation,
Execute
Figure 3-1. V2 ColdFire Core Pipelines
As with all ColdFire cores, the V2 ColdFire core is comprised of two separate pipelines that are decoupled by an instruction buffer.
The instruction fetch pipeline (IFP) is a two-stage pipeline for prefetching instructions. The prefetched instruction stream is then gated into the two-stage operand execution pipeline (OEP), which decodes the
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
3-1
ColdFire Core instruction, fetches the required operands and then executes the required function. Since the IFP and OEP pipelines are decoupled by an instruction buffer which serves as a FIFO queue, the IFP is able to prefetch instructions in advance of their actual use by the OEP thereby minimizing time stalled waiting for instructions.
The V2 ColdFire core pipeline stages include the following:
• Two-stage instruction fetch pipeline (IFP) (plus instruction buffer stage)
— Instruction address generation (IAG)—Calculates the next prefetch address
— Instruction fetch cycle (IC)—Initiates prefetch on the processor’s local bus
— Instruction buffer (IB)—Buffer stage minimizes effects of fetch latency using FIFO queue
• Two-stage operand execution pipeline (OEP)
— Decode and select/operand fetch cycle (DSOC)—Decodes instructions and fetches the required components for effective address calculation, or the operand fetch cycle
— Address generation/execute cycle (AGEX)—Calculates operand address or executes the instruction
When the instruction buffer is empty, opcodes are loaded directly from the IC cycle into the operand execution pipeline. If the buffer is not empty, the IFP stores the contents of the fetch cycle in the IB until it is required by the OEP.
For register-to-register and register-to-memory store operations, the instruction passes through both OEP stages once. For memory-to-register and read-modify-write memory operations, an instruction is effectively staged through the OEP twice: the first time to calculate the effective address and initiate the operand fetch on the processor’s local bus, and the second time to complete the operand reference and perform the required function defined by the instruction.
The resulting pipeline and local bus structure allows the V2 ColdFire core to deliver sustained high performance across a variety of demanding embedded applications.
3.2
Memory Map/Register Description
The following sections describe the processor registers in the user and supervisor programming models.
The appropriate programming model is selected based on the privilege level (user mode or supervisor mode) of the processor as defined by the S bit of the status register (SR).
registers.
The user programming model is the same as the M68000 family microprocessors, consisting of the following registers:
• 16 general-purpose 32-bit registers (D0–D7, A0–A7)
• 32-bit program counter (PC)
• 8-bit condition code register (CCR)
The supervisor programming model is intended to be used only by system control software to implement restricted operating system functions, I/O control, and memory management. All accesses that affect the control features of ColdFire processors are in the supervisor programming model, which consists of registers available in user mode as well as the following control registers:
3-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Core
• 16-bit status register (SR)
• 32-bit supervisor stack pointer (SSP)
• 32-bit vector base register (VBR)
• Two 32-bit memory base address registers (RAMBAR, FLASHBAR)
Table 3-1. ColdFire Core Programming Model
DRc[4–0]
Register
Width
(bits)
Access Reset Value
Written with
MOVEC
Section/Page
Supervisor/User Access Registers
32 R/W 0xCF20_6089 Load: 0x080
Store: 0x180
Load: 0x081
Store: 0x181
Data Register 0 (D0)
Data Register 1 (D1)
Load: 0x082–7
Store: 0x182–7
Data Register 2–7 (D2–D7)
Load: 0x088–8E
Store: 0x188–8E
Address Register 0–6 (A0–A6)
Load: 0x08F
Store: 0x18F
Supervisor/User A7 Stack Pointer (A7)
0x804
0x805
0x806, 0x809,
0x80A, 0x80B
0x807
MAC Status Register (MACSR)
MAC Address Mask Register (MASK)
MAC Accumulators 0–3 (ACC0–3)
0x808
0x80E
0x80F
MAC Accumulator 0,1 Extension Bytes
(ACCext01)
MAC Accumulator 2,3 Extension Bytes
(ACCext23)
Condition Code Register (CCR)
Program Counter (PC)
32
32
32
32
32
32
32
32
32
8
32
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
0x10A0_1070
Undefined
Undefined
Undefined
0x0000_0000
0xFFFF_FFFF
Undefined
Undefined
Undefined
Undefined
Undefined
No
No
No
No
No
No
No
No
No
No
No
Yes
Supervisor Access Only Registers
0x800
0x801
0x80E
0xC04
User/Supervisor A7 Stack Pointer (OTHER_A7)
Vector Base Register (VBR)
Status Register (SR)
Flash Base Address Register (FLASHBAR)
32
32
16
32
R/W
R/W
R/W
R/W
Undefined
0x0000_0000
0x27--
0x0000_0000
No
Yes
No
Yes
0xC05 RAM Base Address Register (RAMBAR) 32 R/W 0x0000_0000 Yes
1
The addresses listed in this column represent the value of the DRc field used when accessing the core registers via the BDM port.
For more information see Chapter 31, “Debug Module.”
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
3-3
ColdFire Core
3.2.1
Data Registers (D0–D7)
Registers D0–D7 are used as data registers for bit (1-bit), byte (8-bit), word (16-bit) and longword (32-bit) operations; they can also be used as index registers.
NOTE
Registers D0 and D1 contain hardware configuration details after reset. See
Section 3.6.14, “Reset Exception ,
” for more details.
DRc[4:0]: Load: 0x080 + n; n = 0-7 (D n )
Store: 0x180 + n; n = 0-7 (D n )
Access: User read/write
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
R
W
Data
Reset
(D2-D7)
– – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –
Reset
(D0, D1)
See Section 3.6.14, “Reset Exception”
Figure 3-2. Data Registers (D0–D7)
3.2.2
Address Registers (A0–A6)
These registers can be used as software stack pointers, index registers, or base address registers; they can also be used for word and longword operations.
DRc[4:0]: Load: 0x088 + n; n = 0–6 (A n )
Store: 0x188 + n; n = 0–6 (A n )
Access: User read/write
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
R
W
Address
Reset – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –
Figure 3-3. Address Registers (A0–A6)
3.2.3
Supervisor/User Stack Pointers (A7 and OTHER_A7)
This ColdFire architecture supports two independent stack pointer (A7) registers—the supervisor stack pointer (SSP) and the user stack pointer (USP). The hardware implementation of these two programmable-visible 32-bit registers does not identify one as the SSP and the other as the USP. Instead, the hardware uses one 32-bit register as the active A7 and the other as OTHER_A7. Thus, the register contents are a function of the processor operation mode, as shown in the following: if SR[S] = 1 then else
A7 = Supervisor Stack Pointer
OTHER_A7 = User Stack Pointer
A7 = User Stack Pointer
OTHER_A7 = Supervisor Stack Pointer
3-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Core
The BDM programming model supports direct reads and writes to A7 and OTHER_A7. It is the responsibility of the external development system to determine, based on the setting of SR[S], the mapping of A7 and OTHER_A7 to the two program-visible definitions (SSP and USP).
To support dual stack pointers, the following two supervisor instructions are included in the ColdFire instruction set architecture to load/store the USP: move.l Ay, USP; move to USP move.l USP, Ax; move from USP
These instructions are described in the ColdFire Family Programmer’s Reference Manual .
NOTE
The USP must be initialized using the mov.l Ay,USP
instruction before any entry into user mode.
DRc[4:0]: Load: 0x08F (A7)
Store: 0x18F (A7)
0x800 (OTHER_A7)
Access: A7: User read/write
OTHER_A7: Supervisor read/write
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
R
Address
W
Reset – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –
Figure 3-4. Stack Pointer Registers (A7 and OTHER_A7)
3.2.4
EMAC Register Description
Multiply-Accumulate Unit (EMAC),”
and include the following registers:
• Four 48-bit accumulator registers partitioned as follows:
— Four 32-bit accumulators (ACC0–ACC3)
— Eight 8-bit accumulator extension bytes (two per accumulator). These are grouped into two
32-bit values for load and store operations (ACCEXT01 and ACCEXT23).
Accumulators and extension bytes can be loaded, copied, and stored, and results from EMAC arithmetic operations generally affect the entire 48-bit destination.
• Eight 8-bit accumulator extensions (two per accumulator), packaged as two 32-bit values for load and store operations (ACCext01 and ACCext23)
• One 16-bit mask register (MASK)
• One 32-bit status register (MACSR) including four indicator bits signaling product or accumulation overflow (one for each accumulator: PAV0–PAV3)
3.2.5
Condition Code Register (CCR)
The CCR is the LSB of the processor status register (SR). Bits 4–0 act as indicator flags for results generated by processor operations. The extend bit (X) is also used as an input operand during multiprecision arithmetic computations.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
3-5
ColdFire Core
DRc[4:0]: LSB of Status Register (SR)
7
0
6
0 R
W
Reset: 0 0
5
0
4
X
3
N
2
Z
0 — —
Figure 3-5. Condition Code Register (CCR)
—
Access: User read-only
1
V
0
C
— —
Field
2
Z
1
V
7–5
4
X
3
N
0
C
Table 3-2. CCR Field Descriptions
Description
Reserved, should be cleared.
Extend condition code bit. Set to the value of the C-bit for arithmetic operations; otherwise not affected or set to a specified result.
Negative condition code bit. Set if the most significant bit of the result is set; otherwise cleared.
Zero condition code bit. Set if the result equals zero; otherwise cleared.
Overflow condition code bit. Set if an arithmetic overflow occurs implying that the result cannot be represented in the operand size; otherwise cleared.
Carry condition code bit. Set if a carry out of the operand msb occurs for an addition, or if a borrow occurs in a subtraction; otherwise cleared
Set to the value of the C bit for arithmetic operations; otherwise not affected.
3.2.6
Program Counter (PC)
The PC contains the address of the currently executing instruction. During instruction execution and exception processing, the processor automatically increments the contents of the PC or places a new value in the PC, as appropriate. The PC is used as a base address for PC-relative operand addressing.
DRc[4:0]: 0x80F (PC) Access: User read/write
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
R
W
Address
Reset – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –
Figure 3-6. Program Counter Register (PC)
3.2.7
Vector Base Register (VBR)
The VBR contains the base address of the exception vector table in memory. To access the vector table, the displacement of an exception vector is added to the value in VBR. The lower 20 bits of the VBR are not implemented by ColdFire processors; they are assumed to be zero, forcing the table to be aligned on a
1 MByte boundary.
3-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Core
DRc[4:0]: 0x801 (VBR) Access: Supervisor read/write
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
R
Base Address
W
Reset 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Figure 3-7. Vector Base Register (VBR)
3.2.8
Status Register (SR)
The SR stores the processor status and includes the CCR, the interrupt priority mask, and other control bits. In supervisor mode, software can access the entire SR. In user mode, only the lower 8 bits are accessible (CCR). The control bits indicate the following states for the processor: trace mode (T bit), supervisor or user mode (S bit), and master or interrupt state (M bit). All defined bits in the SR have read/write access when in supervisor mode. The SR register must be explicitly loaded after reset and before any compare, Bcc, or Scc instructions are executed.
DRc[4:0]: 0x80E (SR)
15
R
W
T
Reset 0
14
0
0
13
S
1
System Byte
12 11
0
M
0
10 9
I
8 7
0
6
0
0 1 1 1 0 0
Figure 3-8. Status Register (SR)
Access: Supervisor read/write
Condition Code Register (CCR)
5
0
4
X
3
N
2
Z
1
V
0
C
0 — — — — —
Field
11
10–8
I
7–5
4–0
CCR
15
T
14
13
S
12
M
Table 3-3. SR Field Descriptions
Description
Trace enable. When set, the processor performs a trace exception after every instruction.
Reserved, should be cleared.
Supervisor/user state. Denotes whether the processor is in supervisor mode (S = 1) or user mode (S = 0).
Master/interrupt state. This bit is cleared by an interrupt exception, and can be set by software during execution of the RTE or move to SR instructions.
Reserved, should be cleared.
Interrupt level mask. Defines the current interrupt level. Interrupt requests are inhibited for all priority levels less than or equal to the current level, except the edge-sensitive level 7 request, which cannot be masked.
Reserved, should be cleared.
Refer to
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
3-7
ColdFire Core
3.2.9
Memory Base Address Registers (RAMBAR, FLASHBAR)
The memory base address registers are used to specify the base address of the internal SRAM and Flash modules and indicate the types of references mapped to each. Each base address register includes a base address, write-protect bit, address space mask bits, and an enable bit. FLASHBAR determines the base address of the on-chip Flash, and RAMBAR determines the base address of the on-chip RAM. For more information, refer to
Section 11.2.1, “SRAM Base Address Register (RAMBAR).”
3.3
Instruction Set Architecture (ISA_A+)
The original ColdFire instruction set architecture (ISA) was derived from the M68000-family opcodes based on extensive analysis of embedded application code. After the initial ColdFire compilers were created, developers identified ISA additions that would enhance both code density and overall performance. Additionally, as users implemented ColdFire-based designs into a wide range of embedded systems, they identified frequently used instruction sequences that could be improved by the creation of new instructions. This observation was especially prevalent in development environments that made use of substantial amounts of assembly language code.
summarizes the instructions added to revision ISA_A to form revision ISA_A+. For more details see the ColdFire Family Programmer’s Reference Manual .
Table 3-4. Instruction Enhancements over Revision ISA_A
Instruction Description
BITREV
BYTEREV
The contents of the destination data register are bit-reversed; that is, new Dn[31] = old Dn[0], new Dn[30] = old Dn[1], ..., new Dn[0] = old Dn[31].
The contents of the destination data register are byte-reversed; that is, new Dn[31:24] = old
Dn[7:0], ..., new Dn[7:0] = old Dn[31:24].
FF1 The data register, Dn, is scanned, beginning from the most-significant bit (Dn[31]) and ending with the least-significant bit (Dn[0]), searching for the first set bit. The data register is then loaded with the offset count from bit 31 where the first set bit appears.
MOVE FROM
USP
USP
→
Destination
MOVE TO USP Source
→
USP
STLDSR Pushes the contents of the status register onto the stack and then reloads the status register with the immediate data value.
3.4
Exception Processing Overview
Exception processing for ColdFire processors is streamlined for performance. The ColdFire processors differ from the M68000 family in that they include:
• A simplified exception vector table
• Reduced relocation capabilities using the vector base register
• A single exception stack frame format
• Use of a single self-aligning stack pointer (for ISA_A implementations only)
3-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Core
All ColdFire processors use an instruction restart exception model. However, Version 2 ColdFire
” for details.
Exception processing includes all actions from the detection of the fault condition to the initiation of fetch for the first handler instruction. Exception processing is comprised of four major steps.
First, the processor makes an internal copy of the SR and then enters supervisor mode by setting the S bit and disabling trace mode by clearing the T bit. The occurrence of an interrupt exception also forces the M bit to be cleared and the interrupt priority mask to be set to the level of the current interrupt request.
Second, the processor determines the exception vector number. For all faults except interrupts, the processor performs this calculation based on the exception type. For interrupts, the processor performs an interrupt-acknowledge (IACK) bus cycle to obtain the vector number from the interrupt controller. The
IACK cycle is mapped to a special acknowledge address space with the interrupt level encoded in the address.
Third, the processor saves the current context by creating an exception stack frame on the system stack.
Processors implementing ISA_A support a single stack pointer in the A7 address register; therefore, there is not notion of separate supervisor and user stack pointer. As a result, the exception stack frame is created at a 0-modulo-4 address on top of the current system stack. For processors implementing all other ISA revisions and supporting 2 stack pointers, the exception stack frame is created at a 0-modulo-4 address on top of the system stack defined by the supervisor stack pointer (SSP). Additionally, the processor uses a simplified fixed-length stack frame for all exceptions. The exception type determines whether the program counter placed in the exception stack frame defines the location of the faulting instruction (fault) or the address of the next instruction to be executed (next).
Fourth, the processor calculates the address of the first instruction of the exception handler. By definition, the exception vector table is aligned on a 1 Mbyte boundary. This instruction address is generated by fetching an exception vector from the table located at the address defined in the vector base register. The index into the exception table is calculated as (4 × vector number). Once the exception vector has been fetched, the contents of the vector determine the address of the first instruction of the desired handler. After the instruction fetch for the first opcode of the handler has been initiated, exception processing terminates and normal instruction processing continues in the handler.
All ColdFire processors support a 1024-byte vector table aligned on any 1 Mbyte address boundary (see
). The table contains 256 exception vectors; the first 64 are defined by Freescale and the remaining 192 are user-defined interrupt vectors.
Table 3-5. Exception Vector Assignments
Vector
Number(s)
2
3
0
1
Vector
Offset (Hex)
0x000
0x004
0x008
0x00C
Stacked
Program
Counter
—
—
Fault
Fault
Assignment
Initial stack pointer
Initial program counter
Access error
Address error
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
3-9
ColdFire Core
Table 3-5. Exception Vector Assignments (continued)
Vector
Number(s)
Vector
Offset (Hex)
Stacked
Program
Counter
Assignment
9
10
11
12
4
5
6–7
8
0x010
0x014
0x018–0x01C
0x020
0x024
0x028
0x02C
0x030
Fault
Fault
—
Fault
Next
Fault
Fault
Next
Illegal instruction
Divide by zero
Reserved
Privilege violation
Trace
Unimplemented line-a opcode
Unimplemented line-f opcode
Debug interrupt
13
14
15–23
24
0x034
0x038
0x03C–0x05C
0x060
—
Fault
—
Next
Reserved
Format error
Reserved
Spurious interrupt
25–31
32–47
48–63
64–255
0x064–0x07C
0x080–0x0BC
0x0C0–0x0FC
0x100–0x3FC
—
Next
—
Next
Reserved
Trap # 0-15 instructions
Reserved
User-defined interrupts
“Fault” refers to the PC of the instruction that caused the exception; “Next” refers to the PC of the next instruction that follows the instruction that caused the fault.
All ColdFire processors inhibit interrupt sampling during the first instruction of all exception handlers.
This allows any handler to effectively disable interrupts, if necessary, by raising the interrupt mask level contained in the status register. In addition, the ISA_A+ architecture includes an instruction (STLDSR) that stores the current interrupt mask level and loads a value into the SR. This instruction is specifically intended for use as the first instruction of an interrupt service routine which services multiple interrupt requests with different interrupt levels. For more details see the ColdFire Family Programmer’s Reference
Manual .
3.5
Exception Stack Frame Definition
the 16-bit format/vector word (F/V) and the 16-bit status register, and the second longword contains the
32-bit program counter address.
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
SSP
→
FORMAT FS[3:2] VECTOR FS[1:0] Status Register
+ 0x4 Program Counter
Figure 3-9. Exception Stack Frame Form
3-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Core
The 16-bit format/vector word contains 3 unique fields:
• A 4-bit format field at the top of the system stack is always written with a value of 4, 5, 6, or 7 by
the processor indicating a two-longword frame format. See Table 3-6
.
Table 3-6. Format Field Encodings
Original SSP @ Time of Exception, Bits 1:0
00
01
10
11
SSP @ 1st
Instruction of
Handler
Original SSP - 8
Original SSP - 9
Original SSP - 10
Original SSP - 11
Format Field
6
7
4
5
• There is a 4-bit fault status field, FS[3:0], at the top of the system stack. This field is defined for
Table 3-7. Fault Status Encodings
FS[3:0]
00 xx
0100
0101
011x
1000
1001
101x
1100
1101
111x
Definition
Reserved
Error on instruction fetch
Reserved
Reserved
Error on operand write
Attempted write to write-protected space
Reserved
Error on operand read
Reserved
Reserved
• The 8-bit vector number, vector[7:0], defines the exception type and is calculated by the processor for all internal faults and represents the value supplied by the interrupt controller in the case of an interrupt. Refer to
.
3.6
Processor Exceptions
3.6.1
Access Error Exception
The exact processor response to an access error depends on the type of memory reference being performed.
For an instruction fetch, the processor postpones the error reporting until the faulted reference is needed by an instruction for execution. Therefore, faults that occur during instruction prefetches that are then followed by a change of instruction flow do not generate an exception. When the processor attempts to
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
3-11
ColdFire Core execute an instruction with a faulted opword and/or extension words, the access error is signaled and the instruction aborted. For this type of exception, the programming model has not been altered by the instruction generating the access error.
If the access error occurs on an operand read, the processor immediately aborts the current instruction’s execution and initiates exception processing. In this situation, any address register updates attributable to the auto-addressing modes, (for example, (An)+,-(An)), have already been performed, so the programming model contains the updated An value. In addition, if an access error occurs during the execution of a
MOVEM instruction loading from memory, any registers already updated before the fault occurs contain the operands from memory.
The V2 ColdFire processor uses an imprecise reporting mechanism for access errors on operand writes.
Because the actual write cycle may be decoupled from the processor’s issuing of the operation, the signaling of an access error appears to be decoupled from the instruction that generated the write.
Accordingly, the PC contained in the exception stack frame merely represents the location in the program when the access error was signaled. All programming model updates associated with the write instruction are completed. The NOP instruction can collect access errors for writes. This instruction delays its execution until all previous operations, including all pending write operations, are complete. If any previous write terminates with an access error, it is guaranteed to be reported on the NOP instruction.
3.6.2
Address Error Exception
Any attempted execution transferring control to an odd instruction address (that is, if bit 0 of the target address is set) results in an address error exception.
Any attempted use of a word-sized index register (Xn.w) or a scale factor of 8 on an indexed effective addressing mode generates an address error as does an attempted execution of a full-format indexed addressing mode.
3.6.3
Illegal Instruction Exception
Any attempted execution of an illegal 16-bit opcode (except for line-A and line-F opcodes) generates an illegal instruction exception (vector 4). Additionally, any attempted execution of any non-MAC line-A and most line-F opcode generates their unique exception types, vector numbers 10 and 11, respectively.
ColdFire cores do not provide illegal instruction detection on the extension words on any instruction, including MOVEC.
3.6.4
Divide-By-Zero
Attempting to divide by zero causes an exception (vector 5, offset = 0x014).
3.6.5
Privilege Violation
The attempted execution of a supervisor mode instruction while in user mode generates a privilege violation exception. See the ColdFire Programmer’s Reference Manual for a list of supervisor-mode instructions.
3-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Core
3.6.6
Trace Exception
To aid in program development, all ColdFire processors provide an instruction-by-instruction tracing capability. While in trace mode, indicated by setting of the T bit in the status register (SR[15] = 1), the completion of an instruction execution (for all but the STOP instruction) signals a trace exception. This functionality allows a debugger to monitor program execution.
The STOP instruction has the following effects:
1. The instruction before the STOP executes and then generates a trace exception. In the exception stack frame, the PC points to the STOP opcode.
2. When the trace handler is exited, the STOP instruction is executed, loading the SR with the immediate operand from the instruction.
3. The processor then generates a trace exception. The PC in the exception stack frame points to the instruction after the STOP, and the SR reflects the value loaded in the previous step.
If the processor is not in trace mode and executes a STOP instruction where the immediate operand sets
SR[T], hardware loads the SR and generates a trace exception. The PC in the exception stack frame points to the instruction after the STOP, and the SR reflects the value loaded in step 2.
Because ColdFire processors do not support any hardware stacking of multiple exceptions, it is the responsibility of the operating system to check for trace mode after processing other exception types. As an example, consider the execution of a TRAP instruction while in trace mode. The processor will initiate the TRAP exception and then pass control to the corresponding handler. If the system requires that a trace exception be processed, it is the responsibility of the TRAP exception handler to check for this condition
(SR[T] in the exception stack frame asserted) and pass control to the trace handler before returning from the original exception.
3.6.7
Unimplemented Line-A Opcode
A line-A opcode is defined when bits 15-12 of the opword are 0b1010. This exception is generated by the attempted execution of an undefined line-A opcode.
3.6.8
Unimplemented Line-F Opcode
A line-F opcode is defined when bits 15-12 of the opword are 0b1111. This exception is generated by attempted execution of an undefined line-F opcode.
3.6.9
Debug Interrupt
This special type of program interrupt is discussed in detail in Chapter 31, “Debug Module.”
This exception is generated in response to a hardware breakpoint register trigger. The processor does not generate an IACK cycle but rather calculates the vector number internally (vector number 12).
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
3-13
ColdFire Core
3.6.10
RTE and Format Error Exception
When an RTE instruction is executed, the processor first examines the 4-bit format field to validate the frame type. For a ColdFire core, any attempted RTE execution where the format is not equal to {4,5,6,7} generates a format error. The exception stack frame for the format error is created without disturbing the original RTE frame and the stacked PC pointing to the RTE instruction.
The selection of the format value provides some limited debug support for porting code from M68000 applications. On M68000 family processors, the SR was located at the top of the stack. On those processors, bit 30 of the longword addressed by the system stack pointer is typically zero. Thus, if an RTE is attempted using this “old” format, it generates a format error on a ColdFire processor.
If the format field defines a valid type, the processor: (1) reloads the SR operand, (2) fetches the second longword operand, (3) adjusts the stack pointer by adding the format value to the auto-incremented address after the fetch of the first longword, and then (4) transfers control to the instruction address defined by the second longword operand within the stack frame.
3.6.11
TRAP Instruction Exception
The TRAP #n instruction always forces an exception as part of its execution and is useful for implementing system calls.
3.6.12
Interrupt Exception
Interrupt exception processing includes interrupt recognition and the fetch of the appropriate vector from the interrupt controller using an IACK cycle. See Chapter 12, “Interrupt Controller Module,” for details on the interrupt controller.
3.6.13
Fault-on-Fault Halt
If a ColdFire processor encounters any type of fault during the exception processing of another fault, the processor immediately halts execution with the catastrophic “fault-on-fault” condition. A reset is required to force the processor to exit this halted state.
3.6.14
Reset Exception
Asserting the reset input signal to the processor causes a reset exception. The reset exception has the highest priority of any exception; it provides for system initialization and recovery from catastrophic failure. Reset also aborts any processing in progress when the reset input is recognized. Processing cannot be recovered.
The reset exception places the processor in the supervisor mode by setting the S bit and disables tracing by clearing the T bit in the SR. This exception also clears the M bit and sets the processor’s interrupt priority mask in the SR to the highest level (level 7). Next, the VBR is initialized to zero (0x00000000).
The control registers specifying the operation of any memories (e.g., cache and/or RAM modules) connected directly to the processor are disabled.
3-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Core
NOTE
Other implementation-specific registers are also affected. Refer to each of the modules in this user’s manual for details on these registers.
Once the processor is granted the bus, it then performs two longword read bus cycles. The first longword at address 0 is loaded into the stack pointer and the second longword at address 4 is loaded into the program counter. After the initial instruction is fetched from memory, program execution begins at the address in the PC. If an access error or address error occurs before the first instruction is executed, the processor enters the fault-on-fault halted state.
ColdFire processors load hardware configuration information into the D0 and D1 general-purpose registers after system reset. The hardware configuration information is loaded immediately after the reset-in signal is negated. This allows an emulator to read out the contents of these registers via BDM to determine the hardware configuration.
Information loaded into D0 defines the processor hardware configuration as shown in
Access: User read-only DRc[4:0]: Load: 0x080 (D0)
Store: 0x180 (D0)
30 29 31
R
W
Reset 1 1 0
28
PF
27
0 1
26
1
25
1
24
1
23
0
22
VER
21
0 1
20
0
19
0
18
REV
17
0 0
16
0
15 14 13 12 11 10
R MAC DIV EMAC FPU MMU 0
W
Reset 0 1 1 0 0 0
9
0
0
8
0
0
7
1
6
0
ISA
5
0
Figure 3-10. D0 Hardware Configuration Info
4
0
3
1
2 1
DEBUG
0 0
0
1
Table 3-8. D0 Hardware Configuration Info Field Description
Field Description
31–24
PF
Processor family. This field is fixed to a hex value of 0xCF indicating a ColdFire core is present.
23–20
VER
ColdFire core version number. Defines the hardware microarchitecture version of the ColdFire core.
0010 V2 ColdFire core (This is the value used for this device.)
0011 V3 ColdFire core
0100 V4 ColdFire core
0101 V5 ColdFire core
Else Reserved for future use.
19–16
REV
Processor revision number. The default is 0b0000.
15
MAC
MAC present.This bit signals if the optional multiply-accumulate (MAC) execution engine is present in the processor core.
0 MAC execute engine not present in core. (This is the value used for this device.)
1 MAC execute engine is present in core.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
3-15
ColdFire Core
Table 3-8. D0 Hardware Configuration Info Field Description (continued)
Field Description
14
DIV
Divide present. This bit signals if the hardware divider (DIV) is present in the processor core.
0 Divide execute engine not present in core.
1 Divide execute engine is present in core. (This is the value used for this device.)
13
EMAC
EMAC present. This bit signals if the optional enhanced multiply-accumulate (EMAC) execution engine is present in the processor core.
0 EMAC execute engine not present in core.
1 EMAC execute engine is present in core.(This is the value used for this device.)
12
FPU
11
MMU
FPU present. This bit signals if the optional floating-point (FPU) execution engine is present in the processor core.
0 FPU execute engine not present in core. (This is the value used for this device.)
1 FPU execute engine is present in core.
MMU present. This bit signals if the optional virtual memory management unit (MMU) is present in the processor core.
0 MMU execute engine not present in core. (This is the value used for this device.)
1 MMU execute engine is present in core.
10–8 Reserved.
7–4
ISA
ISA revision. This 4-bit field defines the instruction set architecture (ISA) revision level implemented in the ColdFire processor core.
0000 ISA_A
0001 ISA_B
0010 ISA_C
1000 ISA_A+ (This is the value used for this device)
Else Reserved
3–0
DEBUG
Debug module revision number. This 4-bit field defines the revision level of the debug module implemented in the
ColdFire processor core.
0000 DEBUG_A
0001 DEBUG_B
0010 DEBUG_C
0011 DEBUG_D
0100 DEBUG_E
1001 DEBUG_B+ (This is the value used for this device)
Else Reserved
Information loaded into D1 defines the local memory hardware configuration as shown in the figure below.
3-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Core
DRc[4:0]: Load: 0x081 (D1)
Store: 0x181 (D1)
R
W
Reset
0
31
CLSZ
30
0
29
CCAS
28
0 1
27 26
CCSZ
25 24
0 0 0 0
23 22 21
FLASHSZ
20 19
0
1 0 1 0 0
Access: User read-only
18
0
0
17
0
0
16
0
0
R
W
Reset 0
15
MBSZ
14
0
13
UCAS
12 11
0
10
0
9
0
8
0
7 6 5
SRAMSZ
4 3
0
0 1 0 0 0 0 0 1 1
Figure 3-11. D1 Hardware Configuration Info
1 0
Table 3-9. D1 Hardware Configuration Information Field Description
Field Description
31–30
CLSZ
29–28
CCAS
27–24
CCSZ
Cache line size. This field is fixed to a hex value of 0x0 indicating a 16-byte cache line size.
Configurable cache associativity.
00 Four-way
01 Direct mapped (This is the value used for this device)
Else Reserved for future use
Configurable cache size. Indicates the amount of instruction/data cache.
0000 No configurable cache (This is the value used for this device)
0001 512B configurable cache
0010 1KB configurable cache
0011 2KB configurable cache
0100 4KB configurable cache
0101 8KB configurable cache
0110 16KB configurable cache
0111 32KB configurable cache
1000 64KB configurable cache
Else Reserved
23–20
FLASHSZ
Flash bank size.
0000-0111 No flash
1000 64KB Flash
1001 128KB Flash
1010 256KB Flash (This is the value used for this device)
1011 512KB Flash
Else Reserved for future use.
19–16
15–14
MBSZ
Reserved
Bus size. Defines the width of the ColdFire master Bus datapath.
00 32-bit system bus datapath (This is the value used for this device)
01 64-bit system bus datapath
Else Reserved
2
0
0
1
0
0
0
0 0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
3-17
ColdFire Core
Table 3-9. D1 Hardware Configuration Information Field Description (continued)
Field Description
13–12
UCAS
11–8
7–4
SRAMSZ
3-0
Unified cache associativity. Defines the unified cache set-associativity.
00 Four-way
01 Direct mapped (This is the value used for this device)
Else Reserved for future use
Reserved.
SRAM bank size.
0000 No SRAM
0001 512 bytes
0010 1 Kbytes
0011 2 Kbytes
0100 4 Kbytes
0101 8 Kbytes
0110 16 Kbytes
0111 32 Kbytes (This is the value used for this device)
1000 64 Kbytes
1001 128 Kbytes
Else Reserved for future use
Reserved
3.7
Instruction Execution Timing
This section presents processor instruction execution times in terms of processor core clock cycles. The number of operand references for each instruction is enclosed in parentheses following the number of processor clock cycles. Each timing entry is presented as C(R/W) where:
• C is the number of processor clock cycles, including all applicable operand fetches and writes, and all internal core cycles required to complete the instruction execution.
• R/W is the number of operand reads (R) and writes (W) required by the instruction. An operation performing a read-modify-write function is denoted as (1/1).
This section includes the assumptions concerning the timing values and the execution time details.
3.7.1
Timing Assumptions
For the timing data presented in this section, the following assumptions apply:
1. The OEP is loaded with the opword and all required extension words at the beginning of each instruction execution. This implies that the OEP does not wait for the IFP to supply opwords and/or extension words.
2. The OEP does not experience any sequence-related pipeline stalls.The most common example of this type of stall involves consecutive store operations, excluding the MOVEM instruction. For all
STORE operations (except MOVEM), certain hardware resources within the processor are marked as “busy” for two clock cycles after the final decode and select/operand fetch cycle (DSOC) of the store instruction. If a subsequent STORE instruction is encountered within this 2-cycle window, it
3-18
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Core is stalled until the resource again becomes available. Thus, the maximum pipeline stall involving consecutive STORE operations is 2 cycles. The MOVEM instruction uses a different set of resources and this stall does not apply.
3. The OEP completes all memory accesses without any stall conditions caused by the memory itself.
Thus, the timing details provided in this section assume that an infinite zero-wait state memory is attached to the processor core.
4. All operand data accesses are aligned on the same byte boundary as the operand size, i.e., 16-bit operands aligned on 0-modulo-2 addresses, 32-bit operands aligned on 0-modulo-4 addresses.
The processor core decomposes misaligned operand references into a series of aligned accesses as shown
Table 3-10. Misaligned Operand References address[1:0]
01 or 11
01 or 11
10
Size
Word
Long
Long
Bus
Operations
Byte, Byte
Byte, Word,
Byte
Word, Word
Additional
C(R/W)
2(1/0) if read
1(0/1) if write
3(2/0) if read
2(0/2) if write
2(1/0) if read
1(0/1) if write
3.7.2
MOVE Instruction Execution Times
lists execution times for MOVE.{B,W} instructions; Table 3-12 lists timings for MOVE.L. For
all tables in this section, the execution time of any instruction using the PC-relative effective addressing modes is the same for the comparable An-relative mode.The nomenclature “xxx.wl” refers to both forms of absolute addressing, xxx.w and xxx.l.
Table 3-11. MOVE Byte and Word Execution Times
Source
Dy
Ay
(Ay)
(Ay)+
-(Ay)
(d16,Ay)
(d8,Ay,Xi*SF) xxx.w
xxx.l
(d16,PC)
Rx
1(0/0)
1(0/0)
3(1/0)
3(1/0)
3(1/0)
3(1/0)
4(1/0)
3(1/0)
3(1/0)
3(1/0)
(Ax)
1(0/1)
1(0/1)
3(1/1)
3(1/1)
3(1/1)
3(1/1)
4(1/1)
3(1/1)
3(1/1)
3(1/1)
(Ax)+
1(0/1)
1(0/1)
3(1/1)
3(1/1)
3(1/1)
3(1/1)
4(1/1)
3(1/1)
3(1/1)
3(1/1)
Destination
-(Ax)
1(0/1)
1(0/1)
3(1/1)
3(1/1)
3(1/1)
3(1/1)
4(1/1))
3(1/1)
3(1/1)
31/1)
(d16,Ax) (d8,Ax,Xi*SF) xxx.wl
1(0/1)
1(0/1)
3(1/1)
3(1/1)
3(1/1)
3(1/1)
—
—
—
3(1/1)
2(0/1)
2(0/1)
4(1/1))
4(1/1))
4(1/1)
—
—
—
—
—
1(0/1)
1(0/1)
3(1/1)
3(1/1)
3(1/1)
—
—
—
—
—
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
3-19
ColdFire Core
Dy
Ay
(Ay)
(Ay)+
-(Ay)
(d16,Ay)
(d8,Ay,Xi*SF) xxx.w
xxx.l
(d16,PC)
(d8,PC,Xi*SF)
#xxx
Source
(d8,PC,Xi*SF)
#xxx
Source
Rx
1(0/0)
1(0/0)
2(1/0)
2(1/0)
2(1/0)
2(1/0)
3(1/0)
2(1/0)
2(1/0)
2(1/0)
3(1/0)
1(0/0)
Table 3-11. MOVE Byte and Word Execution Times (continued)
Destination
Rx
4(1/0)
1(0/0)
(Ax)
4(1/1)
3(0/1)
(Ax)+
4(1/1)
3(0/1)
-(Ax)
4(1/1))
3(0/1)
(d16,Ax) (d8,Ax,Xi*SF) xxx.wl
—
—
—
—
—
—
Table 3-12. MOVE Long Execution Times
Destination
(Ax)
1(0/1)
1(0/1)
2(1/1)
2(1/1)
2(1/1)
2(1/1)
3(1/1)
2(1/1)
2(1/1)
2(1/1)
3(1/1)
2(0/1)
(Ax)+
1(0/1)
1(0/1)
2(1/1)
2(1/1)
2(1/1)
2(1/1)
3(1/1)
2(1/1)
2(1/1)
2(1/1)
3(1/1)
2(0/1)
-(Ax)
1(0/1)
1(0/1)
2(1/1)
2(1/1)
2(1/1)
2(1/1)
3(1/1)
2(1/1)
2(1/1)
2(1/1)
3(1/1)
2(0/1)
(d16,Ax) (d8,Ax,Xi*SF) xxx.wl
1(0/1)
1(0/1)
2(1/1)
2(1/1)
2(1/1)
2(1/1)
—
—
—
2(1/1)
—
—
2(0/1)
2(0/1)
3(1/1)
3(1/1)
3(1/1)
—
—
—
—
—
—
—
1(0/1)
1(0/1)
2(1/1)
2(1/1)
2(1/1)
—
—
—
—
—
—
—
3.8
Standard One Operand Instruction Execution Times
Table 3-13. One Operand Instruction Execution Times
Opcode <EA> bitrev byterev clr.b
clr.w
clr.l
ext.w
ext.l
extb.l
ff1
Dx
Dx
<ea>
<ea>
<ea>
Dx
Dx
Dx
Dx
Rn
1(0/0)
1(0/0)
1(0/0)
1(0/0)
1(0/0)
1(0/0)
1(0/0)
1(0/0)
1(0/0)
(An)
—
—
1(0/1)
1(0/1)
1(0/1)
—
—
—
—
(An)+
—
—
1(0/1)
1(0/1)
1(0/1)
—
—
—
—
Effective Address
-(An)
—
—
1(0/1)
1(0/1)
1(0/1)
—
—
—
—
(d16,An) (d8,An,Xn*SF) xxx.wl
—
—
1(0/1)
1(0/1)
1(0/1)
—
—
—
—
—
—
2(0/1)
2(0/1)
2(0/1)
—
—
—
—
—
—
1(0/1)
1(0/1)
1(0/1)
—
—
—
—
#xxx
—
—
—
—
—
—
—
—
—
3-20
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Core
Opcode <EA> neg.l
negx.l
not.l
scc swap tst.b
tst.w
tst.l
Dx
<ea>
<ea>
<ea>
Dx
Dx
Dx
Dx
Table 3-13. One Operand Instruction Execution Times (continued)
Rn
1(0/0)
1(0/0)
1(0/0)
1(0/0)
1(0/0)
1(0/0)
1(0/0)
1(0/0)
(An)
—
3(1/0)
3(1/0)
2(1/0)
—
—
—
—
(An)+
—
3(1/0)
3(1/0)
2(1/0)
—
—
—
—
Effective Address
-(An)
—
3(1/0)
3(1/0)
2(1/0)
—
—
—
—
(d16,An) (d8,An,Xn*SF) xxx.wl
—
3(1/0)
3(1/0)
2(1/0)
—
—
—
—
—
—
—
—
—
4(1/0)
4(1/0)
3(1/0)
—
3(1/0)
3(1/0)
2(1/0)
—
—
—
—
3.9
Standard Two Operand Instruction Execution Times
Table 3-14. Two Operand Instruction Execution Times
Opcode asl.l
asr.l
bchg bchg bclr bclr bset bset add.l
add.l
addi.l
addq.l
addx.l
and.l
and.l
andi.l
btst btst cmp.l
cmpi.l
<EA>
Rn
<ea>,Rx
Dy,<ea>
1(0/0)
—
#imm,Dx 1(0/0)
#imm,<ea> 1(0/0)
Dy,Dx
<ea>,Rx
Dy,<ea>
#imm,Dx
1(0/0)
1(0/0)
—
1(0/0)
<ea>,Dx
<ea>,Dx
1(0/0)
1(0/0)
Dy,<ea> 2(0/0)
#imm,<ea> 2(0/0)
Dy,<ea> 2(0/0)
#imm,<ea> 2(0/0)
Dy,<ea> 2(0/0)
#imm,<ea> 2(0/0)
Dy,<ea> 2(0/0)
#imm,<ea> 1(0/0)
<ea>,Rx
#imm,Dx
1(0/0)
1(0/0)
(An)
—
—
4(1/1)
4(1/1)
4(1/1)
4(1/1)
4(1/1)
4(1/1)
3(1/0)
3(1/1)
—
3(1/1)
—
3(1/0)
3(1/1)
—
3(1/1)
3(1/1)
3(1/0)
—
(An)+
Effective Address
-(An)
—
—
4(1/1)
4(1/1)
4(1/1)
4(1/1)
41/1)
4(1/1)
3(1/0) 3(1/0)
3(1/1) 3(1/1)
—
3(1/1)
—
3(1/1)
—
3(1/0)
—
3(1/0)
3(1/1) 3(1/1)
— —
3(1/1)
3(1/1)
3(1/0)
—
3(1/1)
3(1/1)
3(1/0)
—
—
—
4(1/1)
4(1/1)
4(1/1)
4(1/1)
4(1/1)
4(1/1)
(d16,An)
(d16,PC)
(d8,An,Xn*SF)
(d8,PC,Xn*SF) xxx.wl
—
—
4(1/1)
4(1/1)
4(1/1)
4(1/1)
4(1/1)
4(1/1)
3(1/0)
3(1/1)
—
3(1/1)
—
3(1/0)
3(1/1)
—
3(1/1)
3(1/1)
3(1/0)
—
—
—
5(1/1)
—
5(1/1)
—
5(1/1)
—
4(1/0)
4(1/1)
—
4(1/1)
—
4(1/0)
4(1/1)
—
4(1/1)
—
4(1/0)
—
—
—
4(1/1)
—
4(1/1)
—
4(1/1)
—
3(1/0)
3(1/1)
—
3(1/1)
—
3(1/0)
3(1/1)
—
3(1/1)
—
3(1/0)
—
#xxx
—
—
—
—
1(0/0)
1(0/0)
—
—
1(0/0)
—
—
—
—
1(0/0)
—
—
—
1(0/0)
1(0/0)
—
#xxx
—
1(0/0)
1(0/0)
1(0/0)
—
—
—
—
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
3-21
ColdFire Core
Table 3-14. Two Operand Instruction Execution Times (continued) lsr.l
moveq or.l
or.l
ori.l
rems.l
remu.l
sub.l
divs.w
divu.w
divs.l
divu.l
eor.l
eori.l
lea lsl.l
sub.l
subi.l
subq.l
subx.l
Opcode
Effective Address
<EA>
Rn (An) (An)+ -(An)
(d16,An)
(d16,PC)
(d8,An,Xn*SF)
(d8,PC,Xn*SF) xxx.wl
<ea>,Dx 20(0/0) 23(1/0) 23(1/0) 23(1/0) 23(1/0)
<ea>,Dx 20(0/0) 23(1/0) 23(1/0) 23(1/0) 23(1/0)
<ea>,Dx
≤
35(0/0)
≤
38(1/0)
≤
38(1/0)
≤
38(1/0)
≤
38(1/0)
<ea>,Dx
≤
35(0/0)
≤
38(1/0)
≤
38(1/0)
≤
38(1/0)
≤
38(1/0)
Dy,<ea> 1(0/0) 3(1/1) 3(1/1) 3(1/1) 3(1/1)
#imm,Dx
<ea>,Ax
<ea>,Dx
<ea>,Dx
1(0/0)
—
1(0/0)
1(0/0)
—
1(0/0)
—
—
—
—
—
—
—
—
—
—
—
1(0/0)
—
—
#imm,Dx
<ea>,Rx
—
1(0/0)
—
3(1/0)
—
3(1/0)
—
3(1/0)
—
3(1/0)
Dy,<ea> — 3(1/1) 3(1/1) 3(1/1) 3(1/1)
#imm,Dx 1(0/0) — — — —
<ea>,Dx
≤
35(0/0)
≤
38(1/0)
≤
38(1/0)
≤
38(1/0)
≤
38(1/0)
<ea>,Dx
≤
35(0/0)
≤
38(1/0)
≤
38(1/0)
≤
38(1/0)
≤
38(1/0)
<ea>,Rx 1(0/0) 3(1/0) 3(1/0) 3(1/0) 3(1/0)
Dy,<ea>
#imm,Dx
—
1(0/0)
#imm,<ea> 1(0/0)
Dy,Dx 1(0/0)
3(1/1)
—
3(1/1)
—
3(1/1)
—
3(1/1)
—
3(1/1)
—
3(1/1)
—
3(1/1)
—
3(1/1)
—
—
—
4(1/0)
4(1/1)
—
—
—
4(1/0)
24(1/0)
24(1/0)
—
—
4(1/1)
—
2(0/0)
—
4(1/1)
—
4(1/1)
—
#xxx
—
—
3(1/0)
3(1/1)
—
—
—
3(1/0)
23(1/0) 20(0/0)
23(1/0) 20(0/0)
—
—
—
—
3(1/1)
—
1(0/0)
—
—
—
—
1(0/0)
3(1/1)
—
3(1/1)
—
—
—
—
—
1(0/0)
1(0/0)
1(0/0)
—
—
—
—
1(0/0)
3.10
Miscellaneous Instruction Execution Times
Table 3-15. Miscellaneous Instruction Execution Times
Opcode <EA>
Rn link.w
move.l
Ay,#imm
Ay,USP move.l
USP,Ax move.w
CCR,Dx
2(0/1)
3(0/0)
3(0/0)
1(0/0) move.w
<ea>,CCR 1(0/0) move.w
SR,Dx 1(0/0) move.w
<ea>,SR movec Ry,Rc movem.l
<ea>,&list
7(0/0)
9(0/1)
—
(An)
—
—
—
—
—
—
—
—
1+n(n/0)
(An)+
—
—
—
—
—
—
—
—
—
Effective Address
-(An)
—
—
—
—
—
—
—
—
—
(d16,An) (d8,An,Xn*SF) xxx.wl
—
—
—
—
—
—
—
—
1+n(n/0)
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
#xxx
—
—
—
—
1(0/0)
—
7(0/0)
2
—
—
3-22
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Core
Table 3-15. Miscellaneous Instruction Execution Times (continued)
Effective Address
Opcode <EA>
Rn (An) (An)+ -(An) (d16,An) (d8,An,Xn*SF) xxx.wl
#xxx movem.l
&list,<ea> nop pea pulse stldsr stop trap trapf trapf.w
trapf.l
unlk wddata
<ea>
#imm
#imm
#imm
Ax
<ea>
—
3(0/0)
—
1(0/0)
—
—
—
1(0/0)
1(0/0)
1(0/0)
2(1/0)
—
1+n(0/n)
—
2(0/1)
—
—
—
—
—
—
—
—
3(1/0)
—
—
—
—
—
—
—
—
—
—
—
3(1/0)
—
—
—
—
—
—
—
—
—
—
—
3(1/0)
1+n(0/n)
—
2(0/1)
—
—
—
—
—
—
—
—
3(1/0)
4
—
—
3(0/1)
—
—
—
—
—
—
—
—
4(1/0)
5
—
—
2(0/1)
—
—
—
—
—
—
—
—
3(1/0) wdebug <ea> — 5(2/0) — — 5(2/0) — —
1 n is the number of registers moved by the MOVEM opcode.
2
If a MOVE.W #imm,SR instruction is executed and imm[13] = 1, the execution time is 1(0/0).
3
The execution time for STOP is the time required until the processor begins sampling continuously for interrupts.
4
PEA execution times are the same for (d16,PC).
5
PEA execution times are the same for (d8,PC,Xn*SF).
—
—
—
—
5(0/1)
3(0/0)
3
15(1/2)
—
—
—
—
3(1/0)
—
3.11
EMAC Instruction Execution Times
Table 3-16. EMAC Instruction Execution Times
Effective Address
Opcode <EA>
Rn (An) (An)+ -(An) (d16,An)
(d8,An,
Xn*SF) xxx.wl
#xxx muls.w
mulu.w
muls.l
mulu.l
mac.w
mac.l
msac.w
msac.l
<ea>y, Dx
<ea>y, Dx
<ea>y, Dx
<ea>y, Dx
Ry, Rx, Raccx
Ry, Rx, Raccx
Ry, Rx, Raccx
Ry, Rx, Raccx
4(0/0) 6(1/0) 6(1/0) 6(1/0) 6(1/0)
4(0/0) 6(1/0) 6(1/0) 6(1/0) 6(1/0)
4(0/0) 6(1/0) 6(1/0) 6(1/0) 6(1/0)
4(0/0) 6(1/0) 6(1/0) 6(1/0) 6(1/0)
1(0/0)
1(0/0)
1(0/0)
1(0/0) mac.w
Ry, Rx, <ea>, Rw, Raccx — mac.l
Ry, Rx, <ea>, Rw, Raccx —
—
—
—
—
—
—
—
—
—
—
—
—
— — — —
2(1/0) 2(1/0) 2(1/0) 2(1/0)
1
2(1/0) 2(1/0) 2(1/0) 2(1/0)
1
71/0) 6(1/0) 4(0/0)
71/0) 6(1/0) 4(0/0)
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
3-23
ColdFire Core
Table 3-16. EMAC Instruction Execution Times (continued)
Effective Address
Opcode <EA>
Rn msac.w
Ry, Rx, <ea>, Rw — msac.l
Ry, Rx, <ea>, Rw, Raccx — move.l
move.l
move.l
move.l
<ea>y, Raccx
Raccy,Raccx
<ea>y, MACSR
<ea>y, Rmask
1(0/0)
1(0/0)
5(0/0)
4(0/0)
(An) (An)+ -(An) (d16,An)
(d8,An,
Xn*SF) xxx.wl
#xxx
2(1/0) 2(1/0) 2(1/0) 2(1/0)
1
2(1/0) 2(1/0) 2(1/0) 2(1/0)
1
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
1(0/0)
—
5(0/0)
4(0/0) move.l
move.l
<ea>y,Raccext01
<ea>y,Raccext23
1(0/0)
1(0/0)
1(0/0)
2
—
—
—
—
—
—
—
—
—
—
—
—
1(0/0)
1(0/0) move.l
move.l
move.l
move.l
Raccx,<ea>x
MACSR,<ea>x
Rmask, <ea>x
Raccext01,<ea.x
1(0/0)
1(0/0)
1(0/0)
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
— move.l
Raccext23,<ea>x 1(0/0) — — — — — —
1
2
Effective address of (d16,PC) not supported
Storing an accumulator requires one additional processor clock cycle when saturation is enabled, or fractional rounding is performed (MACSR[7:4] = 1---, -11-, --11)
—
3.12
Branch Instruction Execution Times
Table 3-17. General Branch Instruction Execution Times
Opcode <EA> bsr jmp jsr rte rts
<ea>
<ea>
Rn
—
—
—
—
—
(An)
—
3(0/0)
3(0/1)
—
—
(An)+
—
—
—
10(2/0)
5(1/0)
—
—
—
—
—
Effective Address
-(An)
(d16,An)
(d16,PC)
(d8,An,Xi*SF)
(d8,PC,Xi*SF)
3(0/1)
3(0/0)
3(0/1)
—
—
—
4(0/0)
4(0/1)
—
— xxx.wl
—
3(0/0)
3(0/1)
—
—
#xxx
—
—
—
—
—
3-24
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Opcode bra bcc
Table 3-18. BRA, Bcc Instruction Execution Times
Forward
Taken
2(0/0)
3(0/0)
Forward
Not Taken
—
1(0/0)
Backward
Taken
2(0/0)
2(0/0)
Backward
Not Taken
—
3(0/0)
ColdFire Core
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
3-25
ColdFire Core
3-26
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 4
Enhanced Multiply-Accumulate Unit (EMAC)
This chapter describes the functionality, microarchitecture, and performance of the enhanced multiply-accumulate (EMAC) unit in the ColdFire family of processors.
4.1
Multiply-Accumulate Unit
The MAC design provides a set of DSP operations that can be used to improve the performance of embedded code while supporting the integer multiply instructions of the baseline ColdFire architecture.
The MAC provides functionality in three related areas:
1. Signed and unsigned integer multiplies
2. Multiply-accumulate operations supporting signed and unsigned integer operands as well as signed, fixed-point, fractional operands
3. Miscellaneous register operations
The ColdFire family supports two MAC implementations with different performance levels and capabilities. The original MAC features a three-stage execution pipeline optimized for 16-bit operands, with a 16x16 multiply array and a single 32-bit accumulator. The EMAC features a four-stage pipeline optimized for 32-bit operands, with a fully pipelined 32 × 32 multiply array and four 48-bit accumulators.
The first ColdFire MAC supported signed and unsigned integer operands and was optimized for 16x16 operations, such as those found in a variety of applications including servo control and image compression.
As ColdFire-based systems proliferated, the desire for more precision on input operands increased. The result was an improved ColdFire MAC with user-programmable control to optionally enable use of fractional input operands.
EMAC improvements target three primary areas:
• Improved performance of 32 × 32 multiply operation.
• Addition of three more accumulators to minimize MAC pipeline stalls caused by exchanges between the accumulator and the pipeline’s general-purpose registers
• A 48-bit accumulation data path to allow the use of a 40-bit product, plus the addition of 8 extension bits to increase the dynamic number range when implementing signal processing algorithms
The three areas of functionality are addressed in detail in following sections. The logic required to support this functionality is contained in a MAC module, as shown in
.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
4-1
Enhanced Multiply-Accumulate Unit (EMAC)
Operand Y
X
Shift 0,1,-1
+/-
Accumulator(s)
Operand X
Figure 4-1. Multiply-Accumulate Functionality Diagram
4.2
Introduction to the MAC
The MAC is an extension of the basic multiplier found in most microprocessors. It is typically implemented in hardware within an architecture and supports rapid execution of signal processing algorithms in fewer cycles than comparable non-MAC architectures. For example, small digital filters can tolerate some variance in an algorithm’s execution time, but larger, more complicated algorithms such as orthogonal transforms may have more demanding speed requirements beyond the scope of any processor architecture and may require full DSP implementation.
To strike a balance between speed, size, and functionality, the ColdFire MAC is optimized for a small set of operations that involve multiplication and cumulative additions. Specifically, the multiplier array is optimized for single-cycle pipelined operations with a possible accumulation after product generation.
This functionality is common in many signal processing applications. The ColdFire core architecture also has been modified to allow an operand to be fetched in parallel with a multiply, increasing overall performance for certain DSP operations.
Consider a typical filtering operation where the filter is defined as in
.
=
∑ a k ( – ) +
∑ b k ( – )
Eqn. 4-1 k = 1 k = 0
Here, the output y(i) is determined by past output values and past input values. This is the general form of an infinite impulse response (IIR) filter. A finite impulse response (FIR) filter can be obtained by setting coefficients a(k) to zero. In either case, the operations involved in computing such a filter are multiplies and product summing. To show this point, reduce the above equation to a simple, four-tap FIR filter, shown
in Equation 4-2 , in which the accumulated sum is a sum of past data values and coefficients.
3
=
∑ b k ( – ) = k = 0
4.3
General Operation
b 0 x i + b 1 ( – ) + b 2 ( – ) + b 3 ( – )
Eqn. 4-2
The MAC speeds execution of ColdFire integer multiply instructions (MULS and MULU) and provides additional functionality for multiply-accumulate operations. By executing MULS and MULU in the MAC,
4-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Enhanced Multiply-Accumulate Unit (EMAC) execution times are minimized and deterministic compared to the 2-bit/cycle algorithm with early termination that the OEP normally uses if no MAC hardware is present.
The added MAC instructions to the ColdFire ISA provide for the multiplication of two numbers, followed by the addition or subtraction of the product to or from the value in an accumulator. Optionally, the product may be shifted left or right by 1 bit before addition or subtraction. Hardware support for saturation arithmetic can be enabled to minimize software overhead when dealing with potential overflow conditions.
Multiply-accumulate operations support 16- or 32-bit input operands of the following formats:
• Signed integers
• Unsigned integers
• Signed, fixed-point, fractional numbers
The EMAC is optimized for single-cycle, pipelined 32 × 32 multiplications. For word- and longword-sized integer input operands, the low-order 40 bits of the product are formed and used with the destination accumulator. For fractional operands, the entire 64-bit product is calculated and either truncated or rounded to the most-significant 40-bit result using the round-to-nearest (even) method before it is combined with the destination accumulator.
For all operations, the resulting 40-bit product is extended to a 48-bit value (using sign-extension for signed integer and fractional operands, zero-fill for unsigned integer operands) before being combined with the 48-bit destination accumulator.
and Figure 4-3 show relative alignment of input operands, the full 64-bit product, the resulting
40-bit product used for accumulation, and 48-bit accumulator formats.
X
OperandY
OperandX
32
32
Product
40
23 “0”
Extended Product
+
Accumulator
8 40
8 40 8
Extension Byte Upper [7:0] Accumulator [31:0] Extension Byte Lower [7:0]
Figure 4-2. Fractional Alignment
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
4-3
Enhanced Multiply-Accumulate Unit (EMAC)
X
OperandY
OperandX
32
32
Product
24 8 32
Extended Product
+
Accumulator
8 8 32
8 8 32
Extension Byte Upper [7:0] Accumulator [31:0]
Extension Byte Lower [7:0]
Figure 4-3. Signed and Unsigned Integer Alignment
Thus, the 48-bit accumulator definition is a function of the EMAC operating mode. Given that each 48-bit accumulator is the concatenation of 16-bit accumulator extension register (ACCext n ) contents and 32-bit
ACC n contents, the specific definitions are as follows: if MACSR[6:5] == 00 /* signed integer mode */
Complete Accumulator[47:0] = {ACCext n [15:0], ACC n [31:0]} if MACSR[6:5] == 01 or 11 /* signed fractional mode */
Complete Accumulator [47:0] = {ACCext n [15:8], ACC n [31:0], ACCext n [7:0]} if MACSR[6:5] == 10 /* unsigned integer mode */
Complete Accumulator[47:0] = {ACCext n [15:0], ACC n [31:0]}
The four accumulators are represented as an array, ACC n , where n selects the register.
Although the multiplier array is implemented in a four-stage pipeline, all arithmetic MAC instructions have an effective issue rate of 1 cycle, regardless of input operand size or type.
All arithmetic operations use register-based input operands, and summed values are stored internally in an accumulator. Thus, an additional move instruction is needed to store data in a general-purpose register.
One new feature found in EMAC instructions is the ability to choose the upper or lower word of a register as a 16-bit input operand. This is useful in filtering operations if one data register is loaded with the input data and another is loaded with the coefficient. Two 16-bit multiply accumulates can be performed without fetching additional operands between instructions by alternating the word choice during the calculations.
The EMAC has four accumulator registers versus the MAC’s single accumulator. The additional registers improve the performance of some algorithms by minimizing pipeline stalls needed to store an accumulator value back to general-purpose registers. Many algorithms require multiple calculations on a given data set.
By applying different accumulators to these calculations, it is often possible to store one accumulator without any stalls while performing operations involving a different destination accumulator.
4-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Enhanced Multiply-Accumulate Unit (EMAC)
The need to move large amounts of data presents an obstacle to obtaining high throughput rates in DSP engines. New and existing ColdFire instructions can accommodate these requirements. A MOVEM instruction can move large blocks of data efficiently by generating line-sized burst references. The ability to simultaneously load an operand from memory into a register and execute a MAC instruction makes some DSP operations such as filtering and convolution more manageable.
The programming model includes a mask register (MASK), which can optionally be used to generate an operand address during MAC + MOVE instructions. The application of this register with auto-increment addressing mode supports efficient implementation of circular data queues for memory operands.
4.4
Memory Map/Register Definition
The following table and sections explain the MAC registers:
Table 4-1. EMAC Memory Map
CPU Space
(Rc)
Register
0x804
0x805
MAC Status Register (MACSR)
MAC Address Mask Register (MASK)
0x806, 0x809,
0x80A, 0x80B
MAC Accumulators 0–3 (ACC0–3)
0x807
0x808
MAC Accumulator 0,1 Extension Bytes (ACCext01)
MAC Accumulator 2,3 Extension Bytes (ACCext23)
Width
(bits)
Access Reset Value Section/Page
32
32
32
R/W
R/W
R/W
0x0000_0000
0xFFFF_FFFF
Undefined
32
32
R/W
R/W
Undefined
Undefined
4.4.1
MAC Status Register (MACSR)
The MAC status register (MACSR) contains a 4-bit operational mode field and condition flags.
Operational mode bits control whether operands are signed or unsigned and whether they are treated as integers or fractions. These bits also control the overflow/saturation mode and the way in which rounding is performed. Negative, zero, and multiple overflow condition flags are also provided.
Address: CPU @ 0x804 Access: User read/write
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8
R 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 PAV n
7 6 5 4 3 2 1 0
N Z V EV
OMC S/U F/I R/T
W
Reset 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Figure 4-4. MAC Status Register (MACSR)
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
4-5
Enhanced Multiply-Accumulate Unit (EMAC)
Field
31–12
11–8
PAV n
7
OMC
6
S/U
5
F/I
4
R/T
3
N
2
Z
Table 4-2. MACSR Field Descriptions
Description
Reserved, should be cleared.
Product/accumulation overflow flags. Contains four flags, one per accumulator, that indicate if past MAC or
MSAC instructions generated an overflow during product calculation or the 48-bit accumulation. When a
MAC or MSAC instruction is executed, the PAV n flag associated with the destination accumulator is used to form the general overflow flag, MACSR[V]. Once set, each flag remains set until V is cleared by a
MOV.L, MACSR instruction or the accumulator is loaded directly.
Overflow/saturation mode. Used to enable or disable saturation mode on overflow. If set, the accumulator is set to the appropriate constant on any operation which overflows the accumulator. Once saturated, the accumulator remains unaffected by any other MAC or MSAC instructions until either the overflow bit is cleared or the accumulator is directly loaded.
Signed/unsigned operations.
In integer mode:
S/U determines whether operations performed are signed or unsigned. It also determines the accumulator value during saturation, if enabled.
0 Signed numbers. On overflow, if OMC is enabled, an accumulator saturates to the most positive
(0x7FFF_FFFF) or the most negative (0x8000_0000) number, depending on both the instruction and the value of the product that overflowed.
1 Unsigned numbers. On overflow, if OMC is enabled, an accumulator saturates to the smallest value
(0x0000_0000) or the largest value (0xFFFF_FFFF), depending on the instruction.
In fractional mode:
S/U controls rounding while storing an accumulator to a general-purpose register.
0 Move accumulator without rounding to a 16-bit value. Accumulator is moved to a general-purpose register as a 32-bit value.
1 The accumulator is rounded to a 16-bit value using the round-to-nearest (even) method when it is moved
to a general-purpose register. See Section 4.4.1.1.1, “Rounding
.” The resulting 16-bit value is stored in the lower word of the destination register. The upper word is zero-filled. The accumulator value is not affected by this rounding procedure.
Fractional/integer mode Determines whether input operands are treated as fractions or integers.
0 Integers can be represented in either signed or unsigned notation, depending on the value of S/U.
1 Fractions are represented in signed, fixed-point, two’s complement notation. Values range from -1 to
1 - 2
-15
for 16-bit fractions and -1 to 1 - 2
-31
for 32-bit fractions. See Section 4.5.2, “Data
."
Round/truncate mode. Controls the rounding procedure for MOV.L ACCx,Rx, or MSAC.L instructions when operating in fractional mode.
0 Truncate. The product’s lsbs are dropped before it is combined with the accumulator. Additionally, when a store accumulator instruction is executed (MOV.L ACCx,Rx), the 8 lsbs of the 48-bit accumulator logic are simply truncated.
1 Round-to-nearest (even). The 64-bit product of two 32-bit, fractional operands is rounded to the nearest
40-bit value. If the low-order 24 bits equal 0x80_0000, the upper 40 bits are rounded to the nearest even
executed (MOV.L ACCx,Rx), the lsbs of the 48-bit accumulator logic are used to round the resulting 16- or 32-bit value. If MACSR[S/U] = 0 and MACSR[R/T] = 1, the low-order 8 bits are used to round the resulting 32-bit fraction. If MACSR[S/U] = 1, the low-order 24 bits are used to round the resulting 16-bit fraction.
Negative. Set if the msb of the result is set, otherwise cleared. N is affected only by MAC, MSAC, and load operations; it is not affected by MULS and MULU instructions.
Zero. Set if the result equals zero, otherwise cleared. This bit is affected only by MAC, MSAC, and load operations; it is not affected by MULS and MULU instructions.
4-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Enhanced Multiply-Accumulate Unit (EMAC)
Field
1
V
0
EV
Table 4-2. MACSR Field Descriptions (continued)
Description
Overflow. Set if an arithmetic overflow occurs on a MAC or MSAC instruction indicating that the result cannot be represented in the limited width of the EMAC. V is set only if a product overflow occurs or the accumulation overflows the 48-bit structure. V is evaluated on each MAC or MSAC operation and uses the appropriate PAV n flag in the next-state V evaluation.
Extension overflow. Signals that the last MAC or MSAC instruction overflowed the 32 lsbs in integer mode or the 40 lsbs in fractional mode of the destination accumulator. However, the result is still accurately represented in the combined 48-bit accumulator structure. Although an overflow has occurred, the correct result, sign, and magnitude are contained in the 48-bit accumulator. Subsequent MAC or MSAC operations may return the accumulator to a valid 32/40-bit result.
summarizes the interaction of the MACSR[S/U,F/I,R/T] control bits.
Table 4-3. Summary of S/U, F/I, and R/T Control Bits
S/U F/I R/T
0
0
0
1
0
1
1
1
1
0
1
1
Operational Modes x Signed, integer
0 Signed, fractional
Truncate on MAC.L and MSAC.L
No round on accumulator stores
1 Signed, fractional
Round on MAC.L and MSAC.L
Round-to-32-bits on accumulator stores x Unsigned, integer
0 Signed, fractional
Truncate on MAC.L and MSAC.L
Round-to-16-bits on accumulator stores
1 Signed, fractional
Round on MAC.L and MSAC.L
Round-to-16-bits on accumulator stores
4.4.1.1
Fractional Operation Mode
This section describes behavior when the fractional mode is used (MACSR[F/I] is set).
4.4.1.1.1
Rounding
When the processor is in fractional mode, there are two operations during which rounding can occur:
1. Execution of a store accumulator instruction (MOV.L ACCx,Rx). The lsbs of the 48-bit accumulator logic are used to round the resulting 16- or 32-bit value. If MACSR[S/U] is cleared, the low-order 8 bits are used to round the resulting 32-bit fraction. If MACSR[S/U] is set, the low-order 24 bits are used to round the resulting 16-bit fraction.
2. Execution of a MAC (or MSAC) instruction with 32-bit operands. If MACSR[R/T] is zero, multiplying two 32-bit numbers creates a 64-bit product that is truncated to the upper 40 bits; otherwise, it is rounded using round-to-nearest (even) method.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
4-7
Enhanced Multiply-Accumulate Unit (EMAC)
To understand the round-to-nearest-even method, consider the following example involving the rounding of a 32-bit number, R0, to a 16-bit number. Using this method, the 32-bit number is rounded to the closest
16-bit number possible. Let the high-order 16 bits of R0 be named R0.U and the low-order 16 bits be R0.L.
• If R0.L is less than 0x8000, the result is truncated to the value of R0.U.
• If R0.L is greater than 0x8000, the upper word is incremented (rounded up).
• If R0.L is 0x8000, R0 is half-way between two 16-bit numbers. In this case, rounding is based on the lsb of R0.U, so the result is always even (lsb = 0).
— If the lsb of R0.U = 1 and R0.L = 0x8000, the number is rounded up.
— If the lsb of R0.U = 0 and R0.L =0x8000, the number is rounded down.
This method minimizes rounding bias and creates as statistically correct an answer as possible.
The rounding algorithm is summarized in the following pseudocode: if R0.L < 0x8000 then Result = R0.U
else if R0.L > 0x8000 then Result = R0.U + 1 else if lsb of R0.U = 0 then Result = R0.U
else Result = R0.U + 1
/* R0.L = 0x8000 */
The round-to-nearest-even technique is also known as convergent rounding.
4.4.1.1.2
Saving and Restoring the EMAC Programming Model
The presence of rounding logic in the output datapath of the EMAC requires that special care be taken during the EMAC’s save/restore process. In particular, any result rounding modes must be disabled during the save/restore process so the exact bit-wise contents of the EMAC registers are accessed. Consider the following memory structure containing the EMAC programming model: int acc0; int acc1; int acc2; int acc3; int accext01; int accext02; int mask; int macsr;
} macState;
The following assembly language routine shows the proper sequence for a correct EMAC state save. This code assumes all Dn and An registers are available for use and the memory location of the state save is defined by A7.
EMAC_state_save: move.l macsr,d7 clr.l d0 move.l d0,macsr move.l acc0,d0 move.l acc1,d1 move.l acc2,d2 move.l acc3,d3
; save the macsr
; zero the register to ...
; disable rounding in the macsr
; save the accumulators
4-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Enhanced Multiply-Accumulate Unit (EMAC) move.l accext01,d4 move.l accext23,d5 move.l mask,d6 movem.l #0x00ff,(a7)
; save the accumulator extensions
; save the address mask
; move the state to memory
The following code performs the EMAC state restore:
EMAC_state_restore: movem.l (a7),#0x00ff move.l #0,macsr move.l d0,acc0 move.l d1,acc1 move.l d2,acc2 move.l d3,acc3 move.l d4,accext01 move.l d5,accext23 move.l d6,mask move.l d7,macsr
; restore the state from memory
; disable rounding in the macsr
; restore the accumulators
; restore the accumulator extensions
; restore the address mask
; restore the macsr
By executing this type of sequence, the exact state of the EMAC programming model can be correctly saved and restored.
4.4.1.1.3
MULS/MULU
MULS and MULU are unaffected by fractional mode operation; operands are still assumed to be integers.
4.4.1.1.4
Scale Factor in MAC or MSAC Instructions
The scale factor is ignored while the MAC is in fractional mode.
4.4.2
Mask Register (MASK)
The 32-bit MASK implements the low-order 16 bits to minimize the alignment complications involved with loading and storing only 16 bits. When the MASK is loaded, the low-order 16 bits of the source operand are actually loaded into the register. When it is stored, the upper 16 bits are all forced to ones.
This register performs a simple AND with the operand address for MAC instructions. That is, the processor calculates the normal operand address and, if enabled, that address is then ANDed with
{0xFFFF, MASK[15:0]} to form the final address. Therefore, with certain MASK bits cleared, the operand address can be constrained to a certain memory region. This is used primarily to implement circular queues in conjunction with the (An)+ addressing mode.
This feature minimizes the addressing support required for filtering, convolution, or any routine that implements a data array as a circular queue. For MAC + MOVE operations, the MASK contents can optionally be included in all memory effective address calculations. The syntax is as follows:
MAC.sz Ry,RxSF,<ea>y&,Rw
The & operator enables the use of MASK and causes bit 5 of the extension word to be set. The exact algorithm for the use of MASK is as follows:
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
4-9
Enhanced Multiply-Accumulate Unit (EMAC) if extension word, bit [5] = 1, the MASK bit, then if <ea> = (An) oa = An & {0xFFFF, MASK} if <ea> = (An)+ oa = An
An = (An + 4) & {0xFFFF, MASK} if <ea> =-(An) oa = (An - 4) & {0xFFFF, MASK}
An = (An - 4) & {0xFFFF, MASK} if <ea> = (d16,An) oa = (An + se_d16) & {0xFFFF0x, MASK}
Here, oa is the calculated operand address and se_d16 is a sign-extended 16-bit displacement. For auto-addressing modes of post-increment and pre-decrement, the calculation of the updated An value is also shown.
Use of the post-increment addressing mode, {(An)+} with the MASK is suggested for circular queue implementations.
Address: CPU @ 0x805 (MASK) Access: User read/write
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
R
Mask
W
Reset 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
Figure 4-5. Mask Register (MASK)
4.4.3
Accumulator Registers (ACC0–3)
The accumulator registers store 32-bits of the MAC operation result. The accumulator extension registers are used to form the entire 48-bit result.
Address: CPU @ 0x806 (ACC0)
CPU @ 0x809 (ACC1)
CPU @ 0x80A (ACC2)
CPU @ 0x80B (ACC3)
Access: User read/write
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
R
W
Accumulator
Reset – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –
Figure 4-6. Accumulator Registers (ACC0–3)
4.4.4
Accumulator Extension Registers (ACCext01, ACCext23)
Each pair of 8-bit accumulator extension fields are concatenated with the corresponding 32-bit accumulator register to form the 48-bit accumulator. For more information on their use, see
4-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Enhanced Multiply-Accumulate Unit (EMAC)
Address: CPU @ 0x807 (ACCext01) Access: User read/write
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
R
W
Accumulator 0 Upper
Extension Byte
Accumulator 0 Lower
Extension Byte
Accumulator 1 Upper
Extension Byte
Accumulator 1 Lower
Extension Byte
Reset – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –
Figure 4-7. Accumulator Extension Register (ACCext01)
Address: CPU @ 0x808 (ACCext23) Access: User read/write
R
W
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
Accumulator 2 Upper
Extension Byte
Accumulator 2 Lower
Extension Byte
Accumulator 3 Upper
Extension Byte
Accumulator 3 Lower
Extension Byte
Reset – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –
Figure 4-8. Accumulator Extension Register (ACCext23)
4.5
EMAC Instruction Set Summary
summarizes EMAC unit instructions.
Table 4-4. EMAC Instruction Summary
Command Mnemonic Description
Multiply Signed
Multiply Unsigned
Multiply Accumulate
MULS <ea>y,Dx
MULU <ea>y,Dx
MAC Ry,RxSF,ACCx
MSAC Ry,RxSF,ACCx
Multiplies two signed operands yielding a signed result
Multiplies two unsigned operands yielding an unsigned result
Multiplies two operands and adds/subtracts the product to/from an accumulator
MAC Ry,Rx,<ea>y,Rw,ACCx
MSAC Ry,Rx,<ea>y,Rw,ACCx
Multiplies two operands and combines the product to an accumulator while loading a register with the memory operand
Multiply Accumulate with Load
Load Accumulator
Store Accumulator
Copy Accumulator
Load MACSR
MOV.L {Ry,#imm},ACCx
MOV.L ACCx,Rx
MOV.L ACCy,ACCx
MOV.L {Ry,#imm},MACSR
Store MACSR MOV.L MACSR,Rx
Store MACSR to CCR MOV.L MACSR,CCR
Load MAC Mask Reg MOV.L {Ry,#imm},MASK
Store MAC Mask Reg MOV.L MASK,Rx
Loads an accumulator with a 32-bit operand
Writes the contents of an accumulator to a CPU register
Copies a 48-bit accumulator
Writes a value to MACSR
Write the contents of MACSR to a CPU register
Write the contents of MACSR to the CCR
Writes a value to the MASK register
Writes the contents of the MASK to a CPU register
Load AccExtensions01 MOV.L {Ry,#imm},ACCext01 Loads the accumulator 0,1 extension bytes with a 32-bit operand
Load AccExtensions23 MOV.L {Ry,#imm},ACCext23 Loads the accumulator 2,3 extension bytes with a 32-bit operand
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
4-11
Enhanced Multiply-Accumulate Unit (EMAC)
Table 4-4. EMAC Instruction Summary (continued)
Command Mnemonic
Store AccExtensions01 MOV.L ACCext01,Rx
Store AccExtensions23 MOV.L ACCext23,Rx
Description
Writes the contents of accumulator 0,1 extension bytes into a
CPU register
Writes the contents of accumulator 2,3 extension bytes into a
CPU register
4.5.1
EMAC Instruction Execution Times
The instruction execution times for the EMAC can be found in Section 3.11, “EMAC Instruction
The EMAC execution pipeline overlaps the AGEX stage of the OEP; that is, the first stage of the EMAC pipeline is the last stage of the basic OEP. EMAC units are designed for sustained, fully-pipelined operation on accumulator load, copy, and multiply-accumulate instructions. However, instructions that store contents of the multiply-accumulate programming model can generate OEP stalls that expose the
EMAC execution pipeline depth, as in the following: mac.w
mov.l
Ry, Rx, Acc0
Acc0, Rz
The mov.l instruction that stores the accumulator to an integer register (Rz) stalls until the program-visible
copy of the accumulator is available. Figure 4-9 shows EMAC timing.
DSOC
AGEX mac mac
Three-cycle regBusy stall mov mov mov
EMAC EX1
EMAC EX2
EMAC EX3
EMAC EX4
Accumulator 0 mac mac mac mac mov old new
Figure 4-9. EMAC-Specific OEP Sequence Stall
In Figure 4-9 , the OEP stalls the store-accumulator instruction for 3 cycles: the depth of the EMAC
pipeline minus 1. The minus 1 factor is needed because the OEP and EMAC pipelines overlap by a cycle, the AGEX stage. As the store-accumulator instruction reaches the AGEX stage where the operation is performed, the just-updated accumulator 0 value is available.
4-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Enhanced Multiply-Accumulate Unit (EMAC)
As with change or use stalls between accumulators and general-purpose registers, introducing intervening instructions that do not reference the busy register can reduce or eliminate sequence-related store-MAC instruction stalls. In fact, a major benefit of the EMAC is the addition of three accumulators to minimize stalls caused by exchanges between the accumulator(s) and the general-purpose registers.
4.5.2
Data Representation
MACSR[S/U,F/I] selects one of the following three modes, where each mode defines a unique operand type:
1. Two’s complement signed integer: In this format, an N-bit operand value lies in the range -2
(N-1)
< operand < 2
(N-1)
- 1. The binary point is right of the lsb.
2. Unsigned integer: In this format, an N-bit operand value lies in the range 0 < operand < 2
N binary point is right of the lsb.
- 1. The
3. Two’s complement, signed fractional: In an N-bit number, the first bit is the sign bit. The remaining bits signify the first N-1 bits after the binary point. Given an N-bit number, its value is given by the equation in
.
a
N-1 a
N-2 a
N-3
... a
2 a
1 a
0
, value = – ( ) +
∑
2
( ) ⋅ ai i = 0
This format can represent numbers in the range -1 < operand < 1 - 2
(N-1)
.
Eqn. 4-3
For words and longwords, the largest negative number that can be represented is -1, whose internal representation is 0x8000 and 0x8000_0000, respectively. The largest positive word is 0x7FFF or (1 - 2
-15 the most positive longword is 0x7FFF_FFFF or (1 - 2
-31
).
);
4.5.3
MAC Opcodes
MAC opcodes are described in the ColdFire Programmer’s Reference Manual .
Note the following:
• Unless otherwise noted, the value of MACSR[N,Z] is based on the result of the final operation that involves the product and the accumulator.
• The overflow (V) flag is handled differently. It is set if the complete product cannot be represented as a 40-bit value (this applies to 32 × 32 integer operations only) or if the combination of the product with an accumulator cannot be represented in the given number of bits. The EMAC design includes an additional product/accumulation overflow bit for each accumulator that are treated as sticky indicators and are used to calculate the V bit on each MAC or MSAC instruction. See
Section 4.4.1, “MAC Status Register (MACSR) .”
• For the MAC design, the assembler syntax of the MAC (multiply and add to accumulator) and
MSAC (multiply and subtract from accumulator) instructions does not include a reference to the single accumulator. For the EMAC, it is expected that assemblers support this syntax and that no explicit reference to an accumulator is interpreted as a reference to ACC0. These assemblers would also support syntaxes where the destination accumulator is explicitly defined.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
4-13
Enhanced Multiply-Accumulate Unit (EMAC)
• The optional 1-bit shift of the product is specified using the notation {<< | >>} SF, where <<1 indicates a left shift and >>1 indicates a right shift. The shift is performed before the product is added to or subtracted from the accumulator. Without this operator, the product is not shifted. If the
EMAC is in fractional mode (MACSR[F/I] is set), SF is ignored and no shift is performed. Because a product can overflow, the following guidelines are implemented:
— For unsigned word and longword operations, a zero is shifted into the product on right shifts.
— For signed, word operations, the sign bit is shifted into the product on right shifts unless the product is zero. For signed, longword operations, the sign bit is shifted into the product unless an overflow occurs or the product is zero, in which case a zero is shifted in.
— For all left shifts, a zero is inserted into the lsb position.
The following pseudocode explains basic MAC or MSAC instruction functionality. This example is presented as a case statement covering the three basic operating modes with signed integers, unsigned integers, and signed fractionals. Throughout this example, a comma-separated list in curly brackets, {}, indicates a concatenation operation.
switch (MACSR[6:5])
{ case 0:
/* MACSR[S/U, F/I] */
/* signed integers */ if (MACSR.OMC == 0 || MACSR.PAVn == 0) then {
MACSR.PAVn = 0
/* select the input operands */ if (sz == word) then {if (U/Ly == 1) then operandY[31:0] = {sign-extended Ry[31], Ry[31:16]} else operandY[31:0] = {sign-extended Ry[15], Ry[15:0]} if (U/Lx == 1) then operandX[31:0] = {sign-extended Rx[31], Rx[31:16]} else operandX[31:0] = {sign-extended Rx[15], Rx[15:0]}
} else {operandY[31:0] = Ry[31:0] operandX[31:0] = Rx[31:0]
}
/* perform the multiply */ product[63:0] = operandY[31:0] * operandX[31:0]
/* check for product overflow */ if ((product[63:39] != 0x0000_00_0) && (product[63:39] != 0xffff_ff_1)) then { /* product overflow */
MACSR.PAVn = 1
MACSR.V = 1 if (inst == MSAC && MACSR.OMC == 1) then if (product[63] == 1) then result[47:0] = 0x0000_7fff_ffff else result[47:0] = 0xffff_8000_0000 else if (MACSR.OMC == 1) then /* overflowed MAC, saturationMode enabled */ if (product[63] == 1) then result[47:0] = 0xffff_8000_0000 else result[47:0] = 0x0000_7fff_ffff
}
4-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Enhanced Multiply-Accumulate Unit (EMAC)
/* sign-extend to 48 bits before performing any scaling */ product[47:40] = {8{product[39]}} /* sign-extend */
/* scale product before combining with accumulator */ switch (SF)
{
/* 2-bit scale factor */ case 0: break;
/* no scaling specified */ case 1: /* SF = “<< 1” */ product[40:0] = {product[39:0], 0} break; case 2: break; case 3:
/* reserved encoding */
/* SF = “>> 1” */ product[39:0] = {product[39], product[39:1]} break;
} if (MACSR.PAVn == 0) then {if (inst == MSAC) then result[47:0] = ACCx[47:0] - product[47:0] else result[47:0] = ACCx[47:0] + product[47:0]
}
/* check for accumulation overflow */ if (accumulationOverflow == 1) then {MACSR.PAVn = 1
MACSR.V = 1 if (MACSR.OMC == 1) then /* accumulation overflow, saturationMode enabled */ if (result[47] == 1) then result[47:0] = 0x0000_7fff_ffff else result[47:0] = 0xffff_8000_0000
}
/* transfer the result to the accumulator */
ACCx[47:0] = result[47:0]
}
MACSR.V = MACSR.PAVn
MACSR.N = ACCx[47] if (ACCx[47:0] == 0x0000_0000_0000) then MACSR.Z = 1 else MACSR.Z = 0 if ((ACCx[47:31] == 0x0000_0) || (ACCx[47:31] == 0xffff_1)) then MACSR.EV = 0 else MACSR.EV = 1 break; case 1,3: /* signed fractionals */ if (MACSR.OMC == 0 || MACSR.PAVn == 0) then {
MACSR.PAVn = 0 if (sz == word) then {if (U/Ly == 1) then operandY[31:0] = {Ry[31:16], 0x0000} else operandY[31:0] = {Ry[15:0], 0x0000} if (U/Lx == 1)
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
4-15
Enhanced Multiply-Accumulate Unit (EMAC) then operandX[31:0] = {Rx[31:16], 0x0000} else operandX[31:0] = {Rx[15:0], 0x0000}
} else {operandY[31:0] = Ry[31:0] operandX[31:0] = Rx[31:0]
}
/* perform the multiply */ product[63:0] = (operandY[31:0] * operandX[31:0]) << 1
/* check for product rounding */ if (MACSR.R/T == 1) then { /* perform convergent rounding */ if (product[23:0] > 0x80_0000) then product[63:24] = product[63:24] + 1 else if ((product[23:0] == 0x80_0000) && (product[24] == 1)) then product[63:24] = product[63:24] + 1
}
/* sign-extend to 48 bits and combine with accumulator */
/* check for the -1 * -1 overflow case */ if ((operandY[31:0] == 0x8000_0000) && (operandX[31:0] == 0x8000_0000)) then product[71:64] = 0x00 else product[71:64] = {8{product[63]}}
/* zero-fill */
/* sign-extend */ if (inst == MSAC) then result[47:0] = ACCx[47:0] - product[71:24] else result[47:0] = ACCx[47:0] + product[71:24]
/* check for accumulation overflow */ if (accumulationOverflow == 1) then {MACSR.PAVn = 1
MACSR.V = 1 if (MACSR.OMC == 1) then /* accumulation overflow, saturationMode enabled */ if (result[47] == 1) then result[47:0] = 0x007f_ffff_ff00 else result[47:0] = 0xff80_0000_0000
}
/* transfer the result to the accumulator */
ACCx[47:0] = result[47:0]
}
MACSR.V = MACSR.PAVn
MACSR.N = ACCx[47] if (ACCx[47:0] == 0x0000_0000_0000) break; then MACSR.Z = 1 else MACSR.Z = 0 if ((ACCx[47:39] == 0x00_0) || (ACCx[47:39] == 0xff_1)) then MACSR.EV = 0 else MACSR.EV = 1 case 2: /* unsigned integers */ if (MACSR.OMC == 0 || MACSR.PAVn == 0) then {
MACSR.PAVn = 0
/* select the input operands */ if (sz == word) then {if (U/Ly == 1) then operandY[31:0] = {0x0000, Ry[31:16]} else operandY[31:0] = {0x0000, Ry[15:0]} if (U/Lx == 1)
4-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Enhanced Multiply-Accumulate Unit (EMAC) then operandX[31:0] = {0x0000, Rx[31:16]} else operandX[31:0] = {0x0000, Rx[15:0]}
} else {operandY[31:0] = Ry[31:0] operandX[31:0] = Rx[31:0]
}
/* perform the multiply */ product[63:0] = operandY[31:0] * operandX[31:0]
/* check for product overflow */ if (product[63:40] != 0x0000_00) then { /* product overflow */
MACSR.PAVn = 1
MACSR.V = 1 if (inst == MSAC && MACSR.OMC == 1) then result[47:0] = 0x0000_0000_0000 else if (MACSR.OMC == 1) then /* overflowed MAC, saturationMode enabled */ result[47:0] = 0xffff_ffff_ffff
}
/* zero-fill to 48 bits before performing any scaling */ product[47:40] = 0 /* zero-fill upper byte */
/* scale product before combining with accumulator */ switch (SF)
{
/* 2-bit scale factor */ case 0: break;
/* no scaling specified */ case 1: /* SF = “<< 1” */ product[40:0] = {product[39:0], 0} break; case 2: break; case 3:
/* reserved encoding */
/* SF = “>> 1” */ product[39:0] = {0, product[39:1]} break;
}
/* combine with accumulator */ if (MACSR.PAVn == 0) then {if (inst == MSAC) then result[47:0] = ACCx[47:0] - product[47:0] else result[47:0] = ACCx[47:0] + product[47:0]
}
/* check for accumulation overflow */ if (accumulationOverflow == 1) then {MACSR.PAVn = 1
MACSR.V = 1 if (inst == MSAC && MACSR.OMC == 1) then result[47:0] = 0x0000_0000_0000 else if (MACSR.OMC == 1) then /* overflowed MAC, saturationMode enabled */
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
4-17
Enhanced Multiply-Accumulate Unit (EMAC) result[47:0] = 0xffff_ffff_ffff
}
}
/* transfer the result to the accumulator */
ACCx[47:0] = result[47:0]
}
MACSR.V = MACSR.PAVn
MACSR.N = ACCx[47] if (ACCx[47:0] == 0x0000_0000_0000) then MACSR.Z = 1 else MACSR.Z = 0 break; if (ACCx[47:32] == 0x0000) then MACSR.EV = 0 else MACSR.EV = 1
4-18
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 5 Cryptographic Acceleration Unit
This chapter describes the Cryptographic Acceleration Unit (CAU) programming model. The CAU is an instruction level coprocessor that is accessed with ColdFire coprocessor instructions (see section XX). The
CAU supports acceleration of the following cryptographic algorithms: DES, 3DES, AES, MD5 and
SHA-1.
NOTE
To enhance readability, this chapter shows the CAU as coprocessor 0.
Future implementations could have the CAU designated as coprocessor 1.
5.1
CAU Registers
The CAU register file consists of eight, 32-bit registers as shown in Table 5-1
. All registers can be read with the coprocessor store instruction (cp0st.l) and written with the coprocessor load instruction (cp0ld.l).
However, only bits 0-1 of the CASR are writable. Bits 2-27 of CASR loads should be 0 for compatibility with future versions of the CAU. The CAU only supports long word accesses and register codes 0x8-0xF are reserved.
Table 5-1. CAU Register File
Code
6
7
4
5
2
3
0
1
Name
CASR
CAA
CA0
CA1
CA2
CA3
CA4
CA5
Description status register accumulator general purpose 0 general purpose 1 general purpose 2 general purpose 3 general purpose 4 general purpose 5
DES AES SHA-1 MD5
--
--
L
R
--
--
C
D
--
--
W0
W1
W2
W3
--
--
--
T
A
B
C
D
E
W
--
-c d
-b
-a
5.1.1
CAU Status Register
The status register (CASR) contains all of the status and configuration for the CAU. It has three defined fields and 26 bits reserved as shown in
Figure 5-1 <f-helvetica><st-bold>Figure 5-1..
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
5-1
Cryptographic Acceleration Unit
31
R
W
RESET: 0
30
VER
29
0 0
28
1
27
0
0
26
0
0
25
0
0
24
0
0
23
0
0
22
0
0
21
0
0
20
0
0
19
0
0
18
0
0
17
0
0
16
0
0
R
W
RESET:
15
0
0
14
0
0
13
0
0
12
0
0
11
0
0
10
0
0
9
0
0
8
0
0
7
0
0
6
0
0
= Read Only or Reserved
Figure 5-1. Status Register (CASR)
IC — Illegal Command
1 = Illegal coprocessor command issued
0 = No illegal commands issued
DPE — DES Parity Error
1 = DES key parity error detected
0 = No error detected
VER — CAU Version
Indicates CAU version; only version 1 is defined.
5
0
0
4
0
0
3
0
0
2
0
0
1 0
DPE IC
0 0
5.2
CAU Operation
The cp0ld.l instruction is used to write to CAU registers and specify CAU operations. Operand 1 of the instruction is the source operand (if any) and the CAU destination register is encoded in the CMD field.
All CAU load instruction commands have an execution time specifier of 0.
The cp0st.l instruction is used to read CAU registers. The CAU source register is encoded in the CMD.
The CAU only supports long word stores. The CAU store instruction command has an execution time specifier of 0.
5.3
CAU Commands
The CAU supports 22 commands as shown in
and described in the following sections, (see
section Section 5.4, “CAU Equate Values
“ for assembly constant definitions). All other encodings are reserved. The IC bit in the CASR will be set if any command is issued that is not defined in the encodings described in this section. A specific illegal command (ILL) is defined to allow for software self checking.
Reserved commands should not be issued to ensure compatibility with future implementations.
NOTE
The value CAx is any CAU register (CASR, CAA, CA0-CA5).
5-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Inst Type cp0ld.l
cp0ld.l
cp0ld.l
cp0ld.l
cp0ld.l
cp0ld.l
cp0ld.l
cp0ld.l
cp0ld.l
cp0ld.l
cp0ld.l
cp0ld.l
cp0st.l
cp0ld.l
cp0ld.l
cp0ld.l
cp0ld.l
cp0ld.l
cp0ld.l
cp0ld.l
cp0ld.l
cp0ld.l
Cryptographic Acceleration Unit
Table 5-2. CAU Commands
Command
Name
MVRA
MVAR
AESS
AESIS
AESC
AESIC
AESR
AESIR
DESR
DESK
CNOP
LDR
STR
ADR
RADR
ADRA
XOR
ROTL
HASH
SHS
MDS
ILL
Description
No Operation
Load Reg
Store Reg
Add
Reverse and Add
Add Reg to Acc
Exclusive Or
Rotate Left
Move Reg to Acc
Move Acc to Reg
AES Sub Bytes
AES Inv Sub Bytes
AES Column Op
AES Inv Column Op
AES Shift Rows
AES Inv Shift Rows
DES Round
DES Key Setup
Hash Function
Secure Hash Shift
Message Digest Shift
Illegal Command
CMD[8:4]
0x08
0x09
0x0A
0x0B
0x0C
0x0D
0x0E
0x0F
0x10
0x11
0x00
0x01
0x02
0x03
0x04
0x05
0x06
0x07
0x12
0x13
0x14
0x1F
CMD[3:0] Operation
CAx
CAx
CAx
CAx
CAx
CAx
0x0
0x0
0x0
CAx
CAx
CAx
CAx
CAx
CAx
CAx
---
Op1 -> CAx
CAx -> Destination
CAx + Op1 -> CAx
CAx + ByteRev(Op1) -> CAx
CAx + CAA -> CAA
CAx ^ Op1 -> CAx
CAx <<< Op1 -> CAx
CAx -> CAA
CAA -> CAx
SubBytes(CAx) -> CAx
InvSubBytes(CAx) -> CAx
MixColumns(CAx)^Op1 -> CAx
InvMixColumns(CAx^Op1) -> CAx
ShiftRows(CA0-CA3) -> CA0-CA3
InvShiftRows(CA0-CA3) -> CA0-CA3
IP FP KS[1:0] DES Round(CA0-CA3)->CA0-CA3
0 0 CP DC DES Key Op(CA0-CA1)->CA0-CA1
Key Parity Error & CP -> CASR[1]
0 HF[2:0]
0x0
Hash Func(CA1-CA3)+CAA->CAA
CAA <<< 5 -> CAA,
CAA->CA0, CA0->CA1,
CA1 <<< 30 -> CA2,
CA2->CA3, CA3->CA4
0x0
0x0
CA3-> CAA, CAA->CA1, CA1->CA2,
CA2->CA3,
0x1->CASR[0]
5.3.1
CNOP - coprocessor no operation
cp0ld.l #CNOP
The CNOP command is the coprocessor cp0nop instruction defined for synchronization.
5.3.2
LDR - load register
cp0ld.l <ea>,#LDR+CAx
The LDR command loads CAx with the source data specified by <ea>.
5.3.3
STR - store register
cp0st.l <ea>,#STR+CAx
The STR command stores the value from CAx to the destination specified by <ea>.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
5-3
Cryptographic Acceleration Unit
5.3.4
ADR - add to register
cp0ld.l <ea>,#ADR+CAx
The ADR command adds the source operand specified by <ea> to CAx and stores the result in CAx.
5.3.5
RADR - reverse and add to register
cp0ld.l <ea>,#RADR+CAx
The RADR command does a byte reverse on the source operand specified by <ea>, adds that value to CAx and stores the result in CAx. An example is shown in
Table 5-3. RADR Command Example
Operand
01020304
CAx Before
A0B0C0D0
CAx After
A4B3C2D1
5.3.6
ADRA - add register to accumulator
cp0ld.l #ADRA+CAx
The ADRA command adds CAx to CAA and stores the result in CAA.
5.3.7
XOR - exclusive or
cp0ld.l <ea>,#XOR+CAx
The XOR command does an exclusive or of the source operand specified by <ea> with CAx and stores the result in CAx.
5.3.8
ROTL - rotate left
cp0ld.l <ea>,#ROTL+CAx
The ROTL rotates the bits of CAx to the left with the result stored back to CAx. The number of bits to rotate is the value specified by <ea> modulo 32.
5.3.9
MVRA - move register to accumulator
cp0ld.l #MVRA+CAx
The MVRA moves the value from the source register CAx to the destination register CAA.
5.3.10
MVAR - move accumulator to register
cp0ld.l #MVAR+CAx
The MVRA command moves the value from source register CAA to the destination register CAx.
5-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Cryptographic Acceleration Unit
5.3.11
AESS - AES substitution
cp0ld.l #AESS+CAx
The AESS command performs the AES byte substitution operation on CAx and stores the result back to
CAx.
5.3.12
AESIS - AES inverse substitution
cp0ld.l #AESIS+CAx
The AESIS command performs the AES inverse byte substitution operation on CAx and stores the result back to CAx.
5.3.13
AESC - AES column operation
cp0ld.l <ea>,#AESC+CAx
The AESC command performs the AES columns operation on the contents of CAx then performs an exclusive or of that result with the source operand specified by <ea> and stores the result in CAx.
5.3.14
AESIC - AES inverse column operation
cp0ld.l <ea>,#AESIC+CAx
The AESIC command performs an exclusive or operation of the source operand specified by <ea> on the contents of CAx followed by the AES inverse mix columns operation on that result and stores the result back in CAx.
5.3.15
AESR - AES shift rows
cp0ld.l #AESR
The AESR command performs the AES shift rows operation on registers CA0, CA1, CA2 and CA3. An example is shown in
.
Table 5-4. AESR Command Example
Register
CA0
CA1
CA2
CA3
Before
01020304
05060708
090A0B0C
0D0E0F00
After
01060B00
050A0F04
090E0308
0D02070C
5.3.16
AESIR - AES inverse shift rows
cp0ld.l #AESIR
The AESR command performs the AES inverse shift rows operation on registers CA0, CA1, CA2 and
CA3. An example is shown in Table 5-5 .
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
5-5
Cryptographic Acceleration Unit
Table 5-5. AESIR Command Example
Register
CA0
CA1
CA2
CA3
Before
01060B00
050A0F04
090E0308
0D02070C
After
01020304
05060708
090A0B0C
0D0E0F00
5.3.17
DESR - DES round
cp0ld.l #DESR+{IP}+{FP}+{KSx}
The DESR command performs a round of the DES algorithm and a key schedule update with the following source and destination designations: CA0=C, CA1=D, CA2=L, CA3=R. If the IP bit is set then the DES initial permutation is performed on CA2 and CA3 before the round operation. If the FP bit is set then the
DES final permutation (inverse initial permutation) is performed on CA2 and CA3 after the round operation. The round operation uses the source values from registers CA0 and CA1 for the key addition operation. The KSx field specifies the shift to use for the key schedule operation used to update the values
.
Table 5-6. Key Shift Function Codes
KSx
Code
2
3
0
1
KSx
Define
KSL1
KSL2
KSR1
KSR2
Shift Function
Left 1
Left 2
Right 1
Right 2
5.3.18
DESK - DES key setup
cp0ld.l #DESK+{CP}+{DC}
The DESK command performs the initial key transformation (permuted choice 1) defined by the DES algorithm on CA0 and CA1 with CA0 containing bits 1-32 of the key and CA1 containing bits 33-64 of the key
1
. If the DC bit is set then no shift operation is performed and the values C
0
and D
0
are stored back to CA0 and CA1 respectively. The DC bit should be set for decrypt operations. If the DC bit is not set then a left shift by 1 is also performed and the values C
1
and D
1
are stored back to CA0 and CA1 respectively.
The DC bit should be 0 for encrypt operations. If the CP bit is set and a key parity error is detected then the DPE bit of the CASR is set, otherwise it is cleared.
5.3.19
HASH - hash function
cp0ld.l #HASH+HFx
The HASH command performs a hashing operation on CA1, CA2 and CA3 and adds that result to the value in CAA and stores the result in CAA. The specific hash function performed is based on the HFx field
1.The DES algorithm numbers the most significant bit of a block as bit1 and the least significant as bit 64.
5-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
HFx
Code
2
3
0
1
4
5
HFx
Define
HFF
HFG
HFH
HFI
HFC
HFM
Cryptographic Acceleration Unit
Table 5-7. Hash Function Codes
Hash Function
MD5 F()
MD5 G()
MD5 H(), SHA Parity()
MD5 I()
SHA Ch()
SHA Maj()
Hash Logic
CA1&CA2 | ~CA1&CA3
CA1&CA3 | CA2&~CA3
CA1^CA2^CA3
CA2^(CA1|~CA3)
CA1&CA2 ^ ~CA1&CA3
CA1&CA2 ^ CA1&CA3 ^ CA2&CA3
5.3.20
SHS - secure hash shift
cp0ld.l #SHS
The SHS command does a set of register to register move and shift operations in parallel that is useful for implementing SHA-1. The following source and destination assignments are made: CAA=CAA<<<5,
CA0=CAA, CA1=CA0, CA2=CA1<<<30, CA3=CA2, CA4=CA3.
5.3.21
MDS - message digest shift
cp0ld.l #MDS
The MDS command does a set of register to register move operations in parallel that is useful for implementing MD5. The following source and destination assignments are made: CAA=CA3,
CA1=CAA, CA2=CA1, CA3=CA2.
5.3.22
ILL - illegal command
cp0ld.l #ILL
The ILL command is a specific illegal command that sets the IC bit in the CASR. All undefined commands are reserved for use in future implementations.
5.4
CAU Equate Values
; CAU Registers (CAx)
.set CASR,0x0
.set CAA,0x1
.set CA0,0x2
.set CA1,0x3
.set CA2,0x4
.set CA3,0x5
.set CA4,0x6
.set CA5,0x7
; CAU Commands
.set CNOP,0x000
.set LDR,0x010
.set STR,0x020
.set ADR,0x030
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
5-7
Cryptographic Acceleration Unit
.set RADR,0x040
.set ADRA,0x050
.set XOR,0x060
.set ROTL,0x070
.set MVRA,0x080
.set MVAR,0x090
.set AESS,0x0A0
.set AESIS,0x0B0
.set AESC,0x0C0
.set AESIC,0x0D0
.set AESR,0x0E0
.set AESIR,0x0F0
.set DESR,0x100
.set DESK,0x110
.set HASH,0x120
.set SHS,0x130
.set MDS,0x140
.set ILL,0x1F0
; DESR Fields
.set IP,0x08 ; initial permutation
.set FP,0x04 ; final permutation
.set KSL1,0x00 ; key schedule left 1 bit
.set KSL2,0x01 ; key schedule left 2 bits
.set KSR1,0x02 ; key schedule right 1 bit
.set KSR2,0x03 ; key schedule right 2 bits
; DESK Field
.set DC,0x01 ; decrypt key schedule
.set CP,0x02 ; check parity
; HASH Functions Codes
.set HFF,0x0 ; MD5 F() CA1&CA2 | ~CA1&CA3
.set HFG,0x1 ; MD5 G() CA1&CA3 | CA2&~CA3
.set HFH,0x2 ; MD5 H(), SHA Parity() CA1^CA2^CA3
.set HFI,0x3 ; MD5 I() CA2^(CA1|~CA3)
.set HFC,0x4 ; SHA Ch() CA1&CA2 ^ ~CA1&CA3
.set HFM,0x5 ; SHA Maj() CA1&CA2 ^ CA1&CA3 ^ CA2&CA3
5-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 6 Random Number Generator Accelerator (RNGA)
A top level block diagram of the RNGA is shown in Figure 6-1 . The module is connects to the IP Bus
defined in SRS version 3.1.1 IP Interface Specification.
Output Register
RNGA Core/Control
Logic
Internal
Bus
Figure 6-1. RNGA Block Diagram
6.1
Overview
The RNGA (Random Number Generator Accelerator) module is a digital integrated circuit capable of generating 32-bit random numbers. It is designed to comply with FIPS-140 standards for randomness and non-determinism. The random bits are generated by clocking shift registers with clocks derived from ring oscillators. The configuration of the shift registers ensures statistically good data (i.e. data that looks random). The oscillators with their unknown frequencies provide the required entropy needed to create random data.
It is important to note that there is no known cryptographic proof showing that this is a secure method of generating random data. In fact, there may be an attack against the random number generator described in this document if its output is used directly in a cryptographic application (the attack is based on the linearity of the internal shift registers). In light of this, it is highly recommended that the random data produced by this module be used as an input seed to a NIST approved (based on DES or SHA-1) or cryptographically secure (RSA Generator or BBS Generator) random number generation algorithm. It is also recommended that other sources of entropy be used along with the RNGA to generate the seed to the pseudorandom algorithm. The more random sources combined to create the seed the better. The following is a list of sources which could be easily combined with the output of this module.
• Current time using highest precision possible.
• Mouse and keyboard motions (or equivalent if being used on a cell phone or PDA).
• Other entropy supplied directly by the user.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 6-1
Random Number Generator Accelerator (RNGA)
6.2
Features
The RNGA includes these distinctive features:
• 32-bit interface
• 32-bit Output Register
• secure mode
• power saving mode
6.3
Modes of Operation
Although the RNGA has several modes, only one is intended for use during normal operation. The other modes were created to aid in verification and testability of the module. The Normal and Oscillator
Frequency Test Modes are entered by setting the appropriate bits in the RNGA Mode register. The Sleep
Mode is entered by setting the appropriate bit in the RNGA Control register. These registers are described in more detail in
Section 6.4.1, “.Register Descriptions
.
• Normal Mode
In this mode the RNGA generates random data. Since this is the default mode of operation, the user is not required to change the mode before requesting random data. This is also the only valid mode when in the secure state. While in this mode, the internal shift registers are driven by internally generated clocks with unknown frequency. Depending on the internal state of the RNGA, these clocks are derived from either the RNGA’s oscillators or a deterministic clock (based on the system clock). For simplicity sake, throughout the rest of this document these clocks will be referred to as the oscillator clocks.
• Secure Mode
In this mode the RNGA is forced into the Normal mode described above. Secure Mode is equivalent to the condition where the RNGA is in Normal mode and is unable to exit that mode. The low power
Sleep Mode can be entered while in the Secure Mode.
• Verification Mode
This mode is provided for verification and testing of the module. While in this mode, the random output is generated by a counter rather than the usual shift registers. The deterministic result allows for easy verification of the surrounding RNGA control logic.
• Oscillator Frequency Test Mode
This mode is provided for testing of the RNGA’s ring oscillators. While in this mode, the shift registers are clocked exclusively by the oscillator clocks (this may not be the case in the Normal
Mode) allowing the oscillator frequency counters (described in Section 6.4.1, “.Register
Descriptions ) to accurately count the pulses received from the oscillator clocks during a given
amount of time.
• Sleep Mode
6-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Memory Map/Register Definition
In this mode the RNGA’s oscillator clocks are shut off. The mode is entered by writing to the Sleep bit in the Control Register. When in this mode, the Output Register will not be loaded.
• Scan Mode
In this mode the RNGA reconfigures much of its untestable logic so it is testable by scan. The mode is entered by driving the block input ipt_mode active. This mode should only be used when scan is used to test the module.
These are high level descriptions only, detailed descriptions of operating modes are contained in later sections.
6.4
Memory Map/Register Definition
The address map for the RNGA is shown in Table 6-1 . The following subsections describe each
addressable register in more detail.
Address
0x00
0x04
0x08
0x0c
0x10
0x14
0x18
0x1c
0x20
0x24
Table 6-1. Module Memory Map
1
Use
Control
Status
Entropy Register
Output Register
Mode
Verification Control
Oscillator Control Counter
Oscillator #1 Counter
Oscillator #2 Counter
Oscillator Counter Status
Access
R/W
R
W
R
R/W
R/W
R/W
R
R
R
= Registers which are inaccessible in secure mode
1
Shaded registers are intended for factory test purposes only. Access to these registers is blocked when the RNGA is in Secure Mode
6.4.1
.
Register Descriptions
This section consists of register descriptions in address order. All RNGA_32IP registers are 32-bit access only.
6.4.1.1
Control Register
Immediately after reset the RNGA begins generating entropy in its internal shift registers. Random data is not pushed to the Output Register until after the Go bit in the Control register is set to a one. After this, a random 32-bit word is pushed to the Output Register every 256 cycles. If the Output Register is full, then
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 6-3
Random Number Generator Accelerator (RNGA) no push will occur. In this way the Output Register will be kept as full as possible. The fields in the Control
Register are defined in
.
R
W
RESET:
REG
ADDR
31:5
0x0000000
0x0000000
4
Sleep
0
3
0
Clear Interrupt
0
2
Interrupt
Mask
0x001F_0000
0
1
High Assurance
0
= Unimplemented or Reserved
0
Go
0
Figure 6-2. Control Register
Go
The Go Bit must be set before the RNGA will begin loading data into the Output Register. This bit is sticky and can only be cleared by a hardware reset or by changing to Secure Mode. Setting the Go bit will not bring the RNGA out of Sleep Mode. Furthermore, the Go bit does not need to be reset after exiting Sleep Mode.
1 = Output Register will be loaded with random data.
0 = Output Register is not loaded with random data.
High Assurance
While this bit is high, the RNGA will notify the SCC if a security violation occurs (i.e. the Output
Register is read while empty). This bit enables Security Violation bit in the Status Register as well as the output rnga_scc_debug port. This bit is sticky and can only be cleared through a hardware reset.
1 = Enable notification of security violations.
0 = Disable notification of security violations.
Interrupt Mask
This bit masks the error interrupt, ipi_error_int.
1 = Interrupt ipi_error_int is masked.
0 = Interrupt ipi_error_int is enabled.
Clear Interrupt
Writing a one to this bit clears the error interrupt as well as the error status bit in the Status Register.
The bit is self clearing.
1 = Clear interrupt ipi_error_int.
0 = Do not clear interrupt ipi_error_int.
Sleep
The RNGA can be placed in low power mode by either asserting the module input ipg_doze, or by setting this Sleep bit. If either of these conditions are met, the oscillators are disabled. De-asserting ipg_doze and resetting the Sleep Bit will cause the RNGA to exit Sleep Mode. The Output Register will not be pushed while the RNGA is in Sleep Mode.
1 = RNGA is in Sleep Mode.
6-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Memory Map/Register Definition
0 = RNGA is not in Sleep Mode.
6.4.1.2
Status Register
RNGA. Only 32-bit reads of this register is supported.
R
31
Osc.
Dead
30:24
0x00
23-16
Output
Register
Size
15:8
Output
Register
Level
7:5
0x0
4
Sleep
3
Error Interrupt
2
Output
Register
Underflow
1
Last Read
Status
0
Security
Violation
W
RESET:
REG
ADDR
0 0x00 0x01 0x00 0x0 0
0x001F_0004
0 0 0 0
= Unimplemented or Reserved
Figure 6-3. Status Register
Security Violation
When enabled by the High Assurance bit in the Control Register, this bit signals that a security violation has occurred. Currently, Output Register underflow is the only condition which is considered a security violation. The bit is sticky and can only be cleared by a hardware reset. The output rnga_scc_debug is driven off this bit.
1 = A security violation has occurred.
0 = No security violations have occurred or the High Assurance bit in the Control Register is not set.
Last Read Status
This bit is always enabled and reflects the status of the most recent read of the Output Register.
1 = Last read was performed while the Output Register was empty (underflow condition).
0 = Last read was performed while the Output Register was not empty.
Output Register Underflow
This bit is always enabled and signals a Output Register underflow condition. The bit is reset by reading the Status Register.
1 = The Output Register has been read while empty since last read of the Status Register.
0 = The Output Register has not been read while empty since last read of the Status Register.
Error Interrupt
This bit is always enabled and signals a Output Register underflow condition. This bit is different from the previous two bits in that it is only reset by a write to the clear interrupt bit in the Control Register.
This bit is not masked by the Interrupt Mask bit of the Control Register.
1 = The Output Register has been read while empty.
0 = The Output Register has not been read while empty.
Sleep
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 6-5
Random Number Generator Accelerator (RNGA)
This bit reflects whether the RNGA is in Sleep mode (i.e. either the Sleep bit in the Control Register is set or the ipg_doze input is asserted). When this bit is a one, the RNGA is in Sleep Mode and the oscillator clocks are inactive. While in this mode, the Output Register will not be loaded and the Output
Register Level will not increase.
1 = The RNGA is in Sleep Mode.
0 = The RNGA is not in Sleep Mode.
Output Register Level
Signals how many random words are currently resident in the Output Register. The bits should be interpreted as an integer (the value 0b00001001 = signals that 0x09 random words are in the Output
Register)
Output Register Size
Signals the actual size of the Output Register. In other words, this is the maximum possible Output
Register Level. The bits should be interpreted as an integer.
Osc. Dead
Indicates that at least one of the shift registers is stuck in its reset state. This information can be used to determine whether the oscillator clocks are operational (i.e. not dead or broken). Verification of the oscillator clocks is achieved through a sequence of commands written to the Verification Control
Register (see
Section 6.4.1.6, “Verification Control Register ). First, the oscillator clocks should be
turned off. Second, the shift registers should be reset. At this point, the oscillator dead bit will assert since the shift registers are in their reset state. Finally, the oscillators should be turned back on. If the oscillator dead bit does not de-assert within a few clock cycles, the oscillators are dead.
1 = At least one oscillator is broken
0 = Both oscillators are operational
Note : This bit is intended to be used for silicon (production) test purposes and the above sequence should be followed to determine if the RNGA oscillators are functional. For simulation purposes, this bit will not be deterministic and should be masked or “don’t cared”.
6.4.1.3
Entropy Register
The Entropy Register is a write-only register which allows the user to insert entropy into the RNGA. This register allows an external user to continually seed the RNGA with externally generated random data.
Although the use of this register is recommended, it is also optional. The Entropy Register can be written at any time during operation and cannot be written too quickly.
6-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Memory Map/Register Definition
Each time the Entropy Register is written, the written value is used to update the internal state of the
RNGA. The update is performed in such a way that the entropy in the RNGA’s internal state is preserved.
Use of the Entropy Register can increase the entropy but never decrease it.
R
W
RESET:
REG
ADDR
31:0
0x00000000
External Entropy
0x00000000
0x001F_0008
= Unimplemented or Reserved
Figure 6-4. Entropy Register
External Entropy
The bits in this field are used to update the internal state of the RNGA.
6.4.1.4
Output Register
The Output Register provides temporary storage for random data generated by the RNGA. The output of the Output Register is accessible through address 0x0c in the RNGA_32IP’s address map. As long as the
Output Register is not empty, a read of this address will return 32 bits of random data. If the Output
Register is read when it is empty, Error Interrupt, Output Register Underflow and Last Read bits in the
Status Register will be set. If the interrupt is enabled in the Control Register ipi_error_int will also be driven active. If the High Assurance bit in the Control Register is set, then rnga_scc_debug will also be
When in Normal Mode, a new random word is pushed into the Output Register every 256 clock cycles (as long as the Output Register is not full). It is very important that the host polls the Status Register to make sure random values are present before reading from the Output Register.
R
W
RESET:
REG
ADDR
31:0
Random Output
0x00000000
0x001F_000C
= Unimplemented or Reserved
Figure 6-5. Output Register
Random Output
32 bits of random data.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 6-7
Random Number Generator Accelerator (RNGA)
6.4.1.5
Mode Register
This register cannot be accessed when the RNGA is in the Secure Mode. The Mode register is used to
configure the RNGA’s mode of operation. Figure 6-6 shows the valid fields in the Mode Register.
R
W
RESET:
REG
ADDR
31:2
0
0x00000000
1
Oscillator Frequency
Test
0
0x001F_0010
0
Verification
0
= Unimplemented or Reserved
Figure 6-6. Mode Register
Verification
When this bit is active, the RNGA is in Verification Mode. While in this mode, the pseudorandom output of the shift registers is ignored and replaced with a counter value that increments every time the output is sampled. This allows the deterministic logic around the shift registers and oscillators to be verified easily.
1 = RNGA is in Verification Mode
0 = RNGA is not in Verification Mode
Oscillator Frequency Test
When this bit is active, the RNGA is in Oscillator Frequency Test Mode. While in this mode, the
RNGA drives the shift registers with the oscillator clocks 100% of the time (while in Normal Mode, the oscillator clocks are occasionally disabled). This allows the approximate frequencies of the oscillators to be calculated accurately.
1 = RNGA is in Oscillator Frequency Test Mode
0 = RNGA is not in Oscillator Frequency Test Mode
6.4.1.6
Verification Control Register
Through use of this register, the RNGA can placed in a deterministic mode allowing verification of the design. A diagram of the register is shown in
.
6-8
R
W
RESET:
REG
ADDR
31:3
0x0000000
0x00000
2
0
Reset Shift
Registers
0
0x14
1
Force System Clock
0
= Unimplemented or Reserved
Figure 6-7. Verification Control Register‘
0
Shift Clock
Off
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Memory Map/Register Definition
Shift Clock Off
The shift registers in the RNGA are clocked by two internally generated clocks. This bit turns these internal clocks on and off.
1 = The shift register clocks are off.
0 = The shift register clocks are on.
Force System Clock
While in Normal Mode, the shift register clocks are derived from free running oscillators with unknown frequency. The shift register clocks can be made deterministic by setting this bit.
1 = The shift registers are driven by a deterministic clock derived from the system clock.
0 = The shift registers are driven by the oscillator based clocks.
Reset Shift Registers
The shift registers in the RNGA are not connected to the asynchronous reset provided to the RNGA.
In order to reset the shift registers, this self clearing bit must be set.
1 = Reset the shift registers.
0 = Do not reset the shift registers.
6.4.1.7
Oscillator Counter Control Register
This register cannot be accessed when the RNGA is in the Secure Mode. The register is used to tell the oscillator test logic how long to count the oscillator clock pulses. The value written to the register is decremented every system clock cycle (ipg_clk) until it reaches zero. When it reaches zero, the Oscillator
Counter registers are frozen (refer to
Section 6.4.1.8, “Oscillator #1 Counter and
given amount of time. In turn, the approximate frequency of each oscillator clock can be computed.
R
W
RESET:
REG
ADDR
31:18
0x0000
0x0000
17:0
Number of clock cycles remaining
Number of clock cycles during which to count the oscillator clock pulses
0x01000
0x08
= Unimplemented or Reserved
Figure 6-8. Oscillator Counter Control Register
6.4.1.8
Oscillator #1 Counter
This register cannot be accessed when the RNGA is in the Secure Mode. The register is used to count the number of oscillator pulses received from Oscillator #1 starting from the time the Oscillator Counter
Control Register is written and ending at the time the Oscillator Counter Control Register reaches zero.
The Oscillator #1 Counter resets when the Oscillator Counter Control Register is written to and stops incrementing when the Oscillator Counter Control Register reaches zero. This register is clocked
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 6-9
Random Number Generator Accelerator (RNGA) asynchronously with respect to the ipg_clk clock domain. In order to avoid unknown values during simulation and meta stable states in silicon, this register should only be read when the Oscillator Counter
Control Register is zero (i.e. when the register is not counting).
31:20
R
W
RESET: 0x000
REG
ADDR
19:0
Number of clock pulses received from Oscillator #1
0x00000
0x0C
= Unimplemented or Reserved
Figure 6-9. Oscillator Counter Control Register
6.4.1.9
Oscillator #2 Counter
This register cannot be accessed when the RNGA is in the Secure Mode. This register behaves like the
Oscillator #1 Counter Register except that it is connected to the second oscillator, and it is accessed at address 0x20. Please refer to
Section 6.4.1.8, “Oscillator #1 Counter .
6.4.1.10
Oscillator Counter Status
This register cannot be accessed unless the RNGA is in the Secure Mode. The bits in this register signal that each oscillator has generated a minimum of 0x400 clock pulses each. After hardware reset, the oscillator counters will count oscillator pulses for 0x1000 system clock cycles. So, immediately after reset, the bits in this register verify that the oscillators are running at least 1/4 the frequency of the system clock.
This information may be useful on the tester where deterministic behavior is required.
R
W
RESET:
REG
ADDR
31:2
0x00000000
0x00000000
0x24
= Unimplemented or Reserved
Figure 6-10. Oscillator Counter Status Register
1
Osc
#2
0
0
Osc
#1
0
Osc #1 Status
Indicates that the clock derived from the first oscillator has toggled at least 0x400 times since the last write of the Oscillator Counter Control Register (the act of resetting the RNGA can be considered such a write since the Oscillator Counter Control Register resets to a non zero value).
1 = Oscillator #1 has toggled 0x400 times
0 = Oscillator #1 has not toggled 0x400 times.
6-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Functional Description
Osc #2 Status
Indicates that the clock derived from the second oscillator has toggled at least 0x400 times since the last write of the Oscillator Counter Control Register (the act of resetting the RNGA can be considered such a write since the Oscillator Counter Control Register resets to a non zero value).
1 = Oscillator #2 has toggled 0x400 times
0 = Oscillator #2 has not toggled 0x400 times.
6.5
Functional Description
The RNGA has three functional areas. They are the Output Register, IF Unit and the RNGA Core/Control
Logic blocks. Each of these can be seen in
. The following sections describe the blocks in more detail.
6.5.1
Output Register
The Output Register provides temporary storage for random data generated by the RNGA Core/Control
Logic block. For simplicity, the output of the Output Register is accessed through the interface at the
address defined in Table 6-1 . The Status register, described in
Section 6.4.1.2, “Status Register
, allows the host to monitor the number of random words in the Output Register through the Output Register Level field. If the host reads from the Output Register when it is empty and the interrupt is enabled, the RNGA will indicate an error. It is very important that the host polls the Status Register to make sure random values are present before reading from the Output Register.
6.5.2
Interface Block
This block translates the interface block signals for slave control of the RNGA. The interface supports
32-bit word aligned accesses only.
6.5.3
RNGA Core/Control Logic Block
This block contains the RNGA’s control logic as well as its core engine used to generate random data. A diagram is shown in
.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 6-11
Random Number Generator Accelerator (RNGA)
Interface control
Core Engine
(Random Number Engine) random data
Control Block
Output Register
Interface
RNGA Core/Control Logic
Figure 6-11. RNGA Logic Block Diagram
6.5.3.1
Control
The Control Block contains the address decoder, all addressable registers, and control state machines for the RNGA. This block is responsible for communication with both the Slave interface and the Output
Register interface. The block also controls the Core Engine to generate random data. The general functionality of the block is as follows. After reset, entropy is generated and stored in the RNGA’s shift registers. After the Go bit is set in the Control register is written, the Output Register is loaded with a random word every 256 cycles. The process of loading the Output Register continues as long as the Output
Register is not full.
6.5.3.2
Core Engine
The Core Engine Block contains the logic used to generate random data. The logic within the Core Engine contains the internal shift registers as well as the logic used to generate the two oscillator based clocks.
This logic is brainless and must be controlled by the Control Block. The Control Block controls how the shift registers are configured as well as when the oscillator clocks are turned on.
6.6
Initialization/Application Information
The intended general operation of the RNGA is as follows:
1. Reset/initialize
2. Write to the Control register and set the Interrupt Mask, High Assurance, and Go bits.
3. Poll Status register for Output Register level
4. Read available random data from Output Register
5. Repeat steps 3 and 4 as needed
6-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Initialization/Application Information
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 6-13
Random Number Generator Accelerator (RNGA)
6-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Chapter 7
Clock Module
7.1
Introduction
The clock module allows the MCF52235 to be configured for one of several clocking methods. Clocking modes include internal phase-locked loop (PLL) clocking with either an external clock reference or an external crystal reference supported by an internal crystal amplifier. The PLL can also be disabled and an external oscillator can be used to clock the device directly. The device always comes out of reset running in external crystal mode (although this mode also supports an external clock source) with the PLL disabled. Once out of reset, it is not possible to change the input clock source, although it is possible to enable the PLL and switch between the PLL clock and the oscillator clock as the source of the system clock.The clock module contains the following:
• Crystal amplifier and oscillator (OSC)
• Phase-locked loop (PLL)
• Reduced frequency divider (RFD)
• Status and control registers
• Control logic
7.2
Features
Features of the clock module include the following:
• 25MHz crystal input
• Provides clock for integrated EPHY
• On-chip PLL can generate core frequencies up to maximum 60MHz operating frequency
7.3
Modes of Operation
The clock module can be operated in normal PLL mode, 1:1 PLL mode or external clock mode (PLL disabled).
7.3.1
Normal PLL Mode
In normal PLL mode, the PLL is fully programmable. It can synthesize frequencies ranging from 4x to 18x the reference frequency and has a post divider capable of reducing this synthesized frequency without disturbing the PLL. The PLL reference can be either a crystal oscillator or an external clock.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
7-1
Clock Module
7.3.2
1:1 PLL Mode
In 1:1 PLL mode, the PLL synthesizes a frequency equal to the external clock input reference frequency.
The post divider is not active.
7.3.3
External Clock Mode
In external clock mode, the PLL is bypassed, and the external clock is applied to EXTAL. The resulting operating frequency is equal to the external clock frequency.
7.4
Low-power Mode Operation
This subsection describes the operation of the clock module in low-power and halted modes of operation.
Low-power modes are described in Chapter 7, “Power Management.”
shows the clock module operation in low-power modes.
Table 7-1. Clock Module Operation in Low-power Modes
Low-power Mode
Wait
Clock Operation
Clocks sent to peripheral modules only
Doze
Stop
Halted
Clocks sent to peripheral modules only
All system clocks disabled
Normal
Mode Exit
Exit not caused by clock module, but normal clocking resumes upon mode exit
Exit not caused by clock module, but normal clocking resumes upon mode exit
Exit not caused by clock module, but clock sources are re-enabled and normal clocking resumes upon mode exit
Exit not caused by clock module
In wait and doze modes, the system clocks to the peripherals are enabled and the clocks to the CPU and memory are stopped. Each module can disable its clock locally at the module level.
In stop mode, all system clocks are disabled. There are several options for enabling or disabling the PLL in stop mode, compromising between stop mode current and wakeup recovery time. The PLL can be disabled in stop mode, but requires a wakeup period before it can relock.
There is also a fast wakeup option for quickly enabling the system clocks during stop recovery. This eliminates the wakeup recovery time but at the risk of sending a potentially unstable clock to the system.
To prevent a non-locked PLL frequency overshoot when using the fast wakeup option, change the RFD divisor to the current RFD value plus one before entering stop mode.
In external clock mode, there are no wakeup periods for oscillator startup or PLL lock.
7.5
Block Diagram
shows a block diagram of the entire clock module. The PLL block in this diagram is expanded
7-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
EXTAL XTAL
EXTERNAL CLOCK
OSC
STOP MODE
STPMD[1:0]
Clock Module
PLLREF
REFERENCE
CLKGEN
CLOCK
LOCEN
PLLSEL
MFD
RSTOUT
PLLMODE
CLKOUT
PLL
LOLRE LOCRE
PLL CLOCK OUT
SCALED PLL CLOCK OUT
INTERNAL CLOCK
LOCKS
LOCK
LOCS
RFD[2:0]
TO RESET
MODULE
CLKOUT
DISCLK
INTERNAL
CLOCKS
STOP MODE
PLLMODE
LOCK
FWKUP
Figure 7-1. Clock Module Block Diagram
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
7-3
Clock Module
LOCEN
REFERENCE
CLOCK
PLLSEL
DISCLK
CLKOUT
RSTOUT STPMD
LOCK
DETECT
LOLRE
PLLMODE
LOCRE
LOSS OF
CLOCK
DETECT
PHASE AND
FREQUENCY
DETECT
CHARGE
PUMP
FILTER VCO
MDF[2:0]
÷ MFD
(4–18)
LOCKS
LOCK
TO RESET
MODULE
LOCS
RFD[2:0]
SCALED PLL
CLOCK OUT
PLL CLOCK
OUT
Figure 7-2. PLL Block Diagram
7.6
Signal Descriptions
information, refer to Chapter 2, “Signal Descriptions.”
Table 7-2. Signal Properties
Name
EXTAL
XTAL
CLKOUT
RSTO
Function
Oscillator or clock input
Oscillator output
System clock output
Reset signal from reset controller
7.6.1
EXTAL
This input is driven by an external clock except when used as a connection to the external crystal when using the internal oscillator.
7-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Clock Module
7.6.2
XTAL
This output is an internal oscillator connection to the external crystal. If CLKMOD0 is driven low during reset, XTAL is sampled to determine clocking mode.
7.6.3
CLKOUT
This output reflects the internal system clock.
7.6.4
RSTO
The RSTO pin is asserted by one of the following:
• Internal system reset signal
• FRCRSTOUT bit in the reset control status register (RCR); see Section 9.5.1, “Reset Control
Register (RCR).”
7.7
Memory Map and Registers
The clock module programming model shown in
Table 7-3 consists of registers that define clock operation
and status as well as additional peripheral power management registers.
Table 7-3. Clock Module Memory Map
IPSBAR
Offset
1
Register
Width
(bits)
Access Reset Value Section/Page
Supervisor Mode Access Only
0x0012_0000 Synthesizer Control Register (SYNCR)
0x0012_0002 Synthesizer Status Register (SYNSR)
0x0012_0006 Low Power Divider Register (LPDR)
0x0012_0008 Clock Control High Register
16
8
8
8
R/W
R
R/W
R/W
0x1002
0x00
0x00
0x00
0x0012_000C Real Time Clock Divide Register
0x0000_000C Peripheral Power Management Register High (PPMRH)
2
32
32
R/W 0x00000000
R/W 0x00000000
0x0000_0008 Peripheral Power Management Register Low (PPMRL)
2
32 R/W 0x00000001
1
2
Addresses not assigned to a register and undefined register bits are reserved for expansion.
See Section 7.2.1, “Peripheral Power Management Registers (PPMRH, PPMRL).”
7.2.1/7-1
7.2.1/7-1
7.7.1
Register Descriptions
This subsection provides a description of the clock module registers.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
7-5
Clock Module
7.7.1.1
Synthesizer Control Register (SYNCR)
IPSBAR
Offset:
0x0012_0000 (SYNCR)
14 15
R
W
Reset
LOLRE
0
MFD2
0
13
MFD1
0
12
MFD0
11
LOCRE
Access: Supervisor read/write
10
RFD2
9
RFD1
8
RFD0
1 0 0 0 0
7 6 5 4 3 2 1 0
R
W
Reset
LOCEN DISCLK FWKUP — — CLKSRC
1
PLLMODE PLLEN
1
0 0 0 0 0 0 1 0
Figure 7-3. Synthesizer Control Register (SYNCR)
1
The reset values of PLLEN and CLKSRC are zero, as the PLL will not be enabled when the device emerges from reset).
Field
15
LOLRE
Table 7-4. SYNCR Field Descriptions
Description
Loss-of-lock reset enable. Determines how the system handles a loss-of-lock indication. When operating in normal mode or 1:1 PLL mode, the PLL must be locked before setting the LOLRE bit. Otherwise, reset is immediately asserted. To prevent an immediate reset, the LOLRE bit must be cleared before writing the
MFD[2:0] bits or entering stop mode with the PLL disabled.
0 No reset on loss of lock
1 Reset on loss of lock
Note: In external clock mode, the LOLRE bit has no effect.
7-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Field
14–12
MFD
11
LOCRE
10–8
RFD
7
LOCEN
6
DISCLK
Clock Module
Table 7-4. SYNCR Field Descriptions (continued)
Description
Multiplication Factor Divider. Contain the binary value of the divider in the PLL feedback loop. The MFD[2:0] value is the multiplication factor applied to the reference frequency. When MFD[2:0] are changed or the PLL is disabled in stop mode, the PLL loses lock. In 1:1 PLL mode, MFD[2:0] are ignored, and the multiplication factor is one.
Note: In external clock mode, the MFD[2:0] bits have no effect.
The following table shows the system frequency multiplier of the reference frequency
1
in normal PLL mode.
MFD[2:0]
000
2
(4x)
001
(6x)
010
(8x)
(3)
011
(10x)
100
(12x)
101
(14x)
110
(16x)
111
(18x)
000 (
÷
1)
001 (
÷
2)
3
010 (
÷
4)
011 (
÷
8)
100 (
÷
16)
101 (
÷
32)
4
2
1
1/2
1/4
1/8
6
3
3/2
3/4
3/8
3/16
8
4
2
1
1/2
1/4
10
5
5/2
5/4
5/8
5/16
12
6
3
3/2
3/4
3/8
14
7
7/2
7/4
7/8
7/16
16
8
4
2
1
1/2
18
9
9/2
9/4
9/8
9/16
110 (
÷
64) 1/16 3/32 1/8 5/32 3/16 7/32 1/4 9/32
111 (
÷
128) 1/32 3/64 1/16 5/64 3/32 7/64 1/8 9/64
1
2
3 f sys
= f ref
x 2(MFD + 2)/2 exp RFD; f ref x 2(MFD + 2)
≤
80 MHz, f sys
≤
MHz
MFD = 000 not valid for f ref
< 3 MHz
Default value out of reset
Loss-of-clock reset enable. Determines how the system handles a loss-of-clock condition. When the
LOCEN bit is clear, LOCRE has no effect. If the LOCS flag in SYNSR indicates a loss-of-clock condition, setting the LOCRE bit causes an immediate reset. To prevent an immediate reset, the LOCRE bit must be cleared before entering stop mode with the PLL disabled.
0 No reset on loss-of-clock
1 Reset on loss-of-clock
Note: In external clock mode, the LOCRE bit has no effect.
Reduced frequency divider field. The binary value written to RFD[2:0] is the PLL frequency divisor; see table in MFD bit description. Changing RFD[2:0] does not affect the PLL or cause a relock delay. Changes in clock frequency are synchronized to the next falling edge of the current system clock. To avoid surpassing the allowable system operating frequency, write to RFD[2:0] only when the LOCK bit is set.
Enables the loss-of-clock function. LOCEN does not affect the loss-of-lock function.
0 Loss-of-clock function disabled
1 Loss-of-clock function enabled
Note: In external clock mode, the LOCEN bit has no effect
.
Disable CLKOUT determines whether CLKOUT is driven. Setting the DISCLK bit holds CLKOUT low.
0 CLKOUT enabled
1 CLKOUT disabled
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
7-7
Clock Module
Field
5
FWKUP
4–3
—
2
CLKSRC
1
PLLMODE
0
PLLEN
Table 7-4. SYNCR Field Descriptions (continued)
Description
Fast wakeup. Determines when the system clocks are enabled during wakeup from stop mode.
0 System clocks enabled only when PLL is locked or operating normally
1 System clocks enabled on wakeup regardless of PLL lock status
Note: When FWKUP = 0, if the PLL or oscillator is enabled and unintentionally lost in stop mode, the PLL wakes up in self-clocked mode or reference clock mode depending on the clock that was lost. In external clock mode, the FWKUP bit has no effect on the wakeup sequence.
Reserved, should be cleared.
Clock Source. Determines whether the PLL output clock or the PLL reference clock is to drive the system clock. This bit is ignored when the PLL is disabled, in which case the PLL reference clock will drive the system clock. Having this separate bit allows the PLL to first be enabled, and then the system clock can be switched to the PLL output clock only after the PLL has locked. When disabling the PLL, the clock can be switched before disabling the PLL so that a smooth transfer is ensured.
0) PLLreference clock (input clock) drives the system clock.
1) PLL output clock drives the system clock (provided the PLL is enabled).
Determines the operating mode of the PLL. This bit should only be changed after reset with the PLL disabled.
0) PLL operates in 1:1 mode
1) PLL operates in normal mode
Enables and disables the PLL. If the PLL is enabled out of reset the chip will not leave the reset state until the PLL is locked and the system clock will be driven by the PLL output clock. Use the CLKSRC control bit to switch the system clock between the PLL output clock and PLL bypass clock once the PLL is enabled.
0) PLL is disabled
1) PLL is enabled
7.7.1.2
Synthesizer Status Register (SYNSR)
The SYNSR is a read-only register that can be read at any time. Writing to the SYNSR has no effect and terminates the cycle normally.
IPSBAR
Offset:
0x0012_0002 (SYNSR) Access: Supervisor read/write
W
Reset:
7
R EXTOSC
1
6
—
5
—
4
LOCKS
3
LOCK
2
LOCS
0 0 See note 1 See note 1 0
Note: 1. See the LOCKS and LOCK bit descriptions.
Figure 7-4. Synthesizer Status Register (SYNSR)
1
—
0
0
—
0
Field
7
EXTOSC
6–5
Table 7-5. SYNSR Field Descriptions
Description
Indicates if an external oscillator is providing the reference clock source
0) Reference clock is not external oscillator
1 Reference clock is external oscillator
Reserved, should be cleared.
7-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Field
4
LOCKS
3
LOCK
2
LOCS
1–0
Clock Module
Table 7-5. SYNSR Field Descriptions (continued)
Description
Sticky indication of PLL lock status.
0 PLL loss of lock since last system reset or MFD change or currently not locked due to exit from STOP with FWKUP set
1 No unintentional PLL loss of lock since last system reset or MFD change
The lock detect function sets the LOCKS bit when the PLL achieves lock after:
• A system reset
• A write to SYNCR that changes the MFD[2:0] bits
When the PLL loses lock, LOCKS is cleared. When the PLL relocks, LOCKS remains cleared until one of the two listed events occurs.
In stop mode, if the PLL is intentionally disabled, then the LOCKS bit reflects the value prior to entering stop mode. However, if FWKUP is set, then LOCKS is cleared until the PLL regains lock. Once lock is regained, the LOCKS bit reflects the value prior to entering stop mode. Furthermore, reading the LOCKS bit at the same time that the PLL loses lock does not return the current loss of lock condition.
In external clock mode, LOCKS remains cleared after reset. In normal PLL mode and 1:1 PLL mode,
LOCKS is set after reset.
Set when the PLL is locked. PLL lock occurs when the synthesized frequency is within approximately 0.75% of the programmed frequency. The PLL loses lock when a frequency deviation of greater than approximately
1.5% occurs. Reading the LOCK flag at the same time that the PLL loses lock or acquires lock does not return the current condition of the PLL. The power-on reset circuit uses the LOCK bit as a condition for releasing reset.
If operating in external clock mode, LOCK remains cleared after reset.
0 PLL not locked
1 PLL locked
Sticky indication of whether a loss-of-clock condition has occurred at any time since exiting reset in normal
PLL and 1:1 PLL modes.
• LOCS = 0 when the system clocks are operating normally.
• LOCS = 1 when system clocks have failed due to a reference failure or PLL failure.
After entering stop mode with FWKUP set and the PLL and oscillator intentionally disabled
(STPMD[1:0] = 11), the PLL exits stop mode in the SCM while the oscillator starts up. During this time,
LOCS is temporarily set regardless of LOCEN. It is cleared once the oscillator comes up and the PLL is attempting to lock.
If a read of the LOCS flag and a loss-of-clock condition occur simultaneously, the flag does not reflect the current loss-of-clock condition.
A loss-of-clock condition can be detected only if LOCEN = 1 or the oscillator has not yet returned from exit from stop mode with FWKUP = 1.
0 Loss-of-clock not detected since exiting reset
1 Loss-of-clock detected since exiting reset or oscillator not yet recovered from exit from stop mode with
FWKUP = 1
Note: The LOCS flag is always 0 in external clock mode.
Reserved, should be cleared.
7.7.1.3
Low Power Divider Register (LPDR)
The low power divider is a 4-bit field that divides down the system clock (regardless if the reference clock or PLL clock is driving the system clock) by a factor of 2 n
(where n is a number from 0 to 15 represented by the 4 bit field). The clock change takes effect with the next rising edge of the system clock.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
7-9
Clock Module
IPSBAR
Offset:
0x0012_0007 (LPDR)
7
— R
W
Reset: 0
6
—
5
—
4
—
3
LPD3
0 0 0 0
Figure 7-5. Low Power Divider Register
2
Access: Supervisor read/write
LPD2
0
1
LPD1
0
0
LPD0
0
7.7.1.4
Clock Control High Register (CCHR)
The Pre-Divider Factor divides down the PLL input clock by 1 (PFD[2:0] = 000) to 8 (PFD[2:0] =111).
This allows an external oscillator or crystal of more than 10 MHz to be used with the PLL. The division factor should be set to generate an input clock for the PLL above 1 MHz and below 10 MHz. When
PFD[2:0] are changed or the PLL is disabled in stop mode, the PLL loses lock. In 1:1 PLL mode, PFD[2:0] are ignored, and the pre-division factor is one.
IPSBAR
Offset:
0x0012_0008 (PFD)
7
— R
W
Reset: 0
6
—
0
5
—
4
—
3
—
0 0 0
Figure 7-6. Clock Control High Register
2
PFD2
0
Access: Supervisor read/write
1
PFD1
0
0
PFD0
0
7.7.1.5
Real Time Clock Divide Register (RTCDR)
The Real Time Clock Divide Factor is a 32 bit read/write register that divides down the oscillator clock to a 1 Hz clock for the Real Time Clock module. If this register is programmed with zero then the clock to the Real Time Clock module is disabled, otherwise the oscillator clock is divided by one more than the value written to the register field (between 2 and 4,294,967,296).
7-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Clock Module
IPSBAR
Offset:
0x0011_000C (RTCDF)
31 30
R
W
Reset 0 0
23 22
R
W
Reset 0
15
0
14
R
W
Reset 0
7
0
6
R
W
Reset 0 0
29
0
13
0
21
0
5
28
RTCDF (31:24)
27
0
20
RTCDF (23:16)
0
19
0
12
RTCDF 15:8)
0
11
0
4
RTCDF (7:0)
0
3
0 0 0
Figure 7-7. Real Time Clock Divide Register
26
0
18
0
2
0
0
10
Access: Supervisor read/write
25
0
9
0
17
0
1
0
24
0
8
0
16
0
0
0
7.8
Functional Description
This section provides a functional description of the clock module.
7.8.1
Clock Operation During Reset
The PLL is always disabled as the part emerges from Reset, with a default configuration of external crystal mode (although this mode also supports an external clock source). Once out of reset, it is not possible to change the input clock source, although it is possible to enable the PLL and switch between the PLL clock and the oscillator clock as the source of the system clock.
7.8.2
System Clock Generation
In normal PLL clock mode, the default system frequency is two times the reference frequency after reset.
The RFD[2:0] and MFD[2:0] bits in the SYNCR select the frequency multiplier. The LPD[3:0] field in the
LPCR register provides additional settings for dividing down the system clock (including when the PLL is disabled) for low power operation.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
7-11
Clock Module
When programming the PLL, do not exceed the maximum system clock frequency listed in the electrical specifications. Use this procedure to accommodate the frequency overshoot that occurs when the MFD bits are changed:
1. Determine the appropriate value for the MFD and RFD fields in the SYNCR. The amount of jitter in the system clocks can be minimized by selecting the maximum MFD factor that can be paired with an RFD factor to provide the required frequency.
2. Write a value of 1 + RFD (from step 1) to the RFD field of the SYNCR.
3. Write the MFD value from step 1 to the SYNCR.
4. Monitor the LOCK flag in SYNSR. When the PLL achieves lock, write the RFD value from step
1 to the RFD field of the SYNCR. This changes the system clocks frequency to the required frequency.
NOTE
Keep the maximum system clock frequency below the limit given in the electrical characteristics.
7.8.3
PLL Operation
In PLL mode, the PLL synthesizes the system clocks. The PLL can multiply the reference clock frequency by 4x to 18x, provided that the system clock frequency remains within the range listed in electrical specifications. For example, if the reference frequency is 2 MHz, the PLL can synthesize frequencies of
8 MHz to 36 MHz. In addition, the RFD can reduce the system frequency by dividing the output of the
PLL. The RFD is not in the feedback loop of the PLL, so changing the RFD divisor does not affect PLL operation.
shows the external support circuitry for the crystal oscillator with example component values.
Actual component values depend on crystal specifications.
The following subsections describe each major block of the PLL. Refer to Figure 7-8 to see how these
functional sub-blocks interact.
C1
C2
25-MHz CRYSTAL CONFIGURATION
C1 = 15 pF, C2 = 39pF
RF = 1 M
Ω
RS = 470
Ω
R1 = 10 M
Ω
R1
V
SSPLL
EXTAL XTAL V
SSPLL
ON-CHIP
RF
RS
7-12
Figure 7-8. Crystal Oscillator Example
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Clock Module
7.8.3.1
Phase and Frequency Detector (PFD)
The PFD is a dual-latch phase-frequency detector. It compares both the phase and frequency of the reference and feedback clocks. The reference clock comes from either the crystal oscillator or an external clock source.
The feedback clock comes from one of the following:
• CLKOUT in 1:1 PLL mode
• VCO output divided by two if CLKOUT is disabled in 1:1 PLL mode
• VCO output divided by the MFD in normal PLL mode
When the frequency of the feedback clock equals the frequency of the reference clock, the PLL is frequency-locked. If the falling edge of the feedback clock lags the falling edge of the reference clock, the
PFD pulses the UP signal. If the falling edge of the feedback clock leads the falling edge of the reference clock, the PFD pulses the DOWN signal. The width of these pulses relative to the reference clock depends on how much the two clocks lead or lag each other. Once phase lock is achieved, the PFD continues to pulse the UP and DOWN signals for very short durations during each reference clock cycle. These short pulses continually update the PLL and prevent the frequency drift phenomenon known as dead-banding.
7.8.3.2
Charge Pump/Loop Filter
In 1:1 PLL mode, the charge pump uses a fixed current. In normal mode the current magnitude of the charge pump varies with the MFD as shown in
.
Table 7-6. Charge Pump Current and MFD in Normal Mode Operation
Charge Pump Current
1x
2x
4x
MFD
0
≤
MFD < 2
2
≤
MFD < 6
6
≤
MFD
The UP and DOWN signals from the PFD control whether the charge pump applies or removes charge, respectively, from the loop filter. The filter is integrated on the chip.
7.8.3.3
Voltage Control Output (VCO)
The voltage across the loop filter controls the frequency of the VCO output. The frequency-to-voltage relationship (VCO gain) is positive, and the output frequency is four times the target system frequency.
7.8.3.4
Multiplication Factor Divider (MFD)
When the PLL is not in 1:1 PLL mode, the MFD divides the output of the VCO and feeds it back to the
PFD. The PFD controls the VCO frequency via the charge pump and loop filter such that the reference and feedback clocks have the same frequency and phase. Thus, the frequency of the input to the MFD, which is also the output of the VCO, is the reference frequency multiplied by the same amount that the MFD divides by. For example, if the MFD divides the VCO frequency by six, the PLL is frequency locked when the VCO frequency is six times the reference frequency. The presence of the MFD in the loop allows the
PLL to perform frequency multiplication, or synthesis.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
7-13
Clock Module
In 1:1 PLL mode, the MFD is bypassed, and the effective multiplication factor is one.
7.8.3.5
PLL Lock Detection
The lock detect logic monitors the reference frequency and the PLL feedback frequency to determine when frequency lock is achieved. Phase lock is inferred by the frequency relationship, but is not guaranteed. The
LOCK flag in the SYNSR reflects the PLL lock status. A sticky lock flag, LOCKS, is also provided.
The lock detect function uses two counters: one is clocked by the reference, and the other is clocked by the PLL feedback. When the reference counter has counted N cycles, its count is compared to that of the feedback counter. If the feedback counter has also counted N cycles, the process is repeated for N + K counts. Then, if the two counters still match, the lock criteria is relaxed by 1/2 and the system is notified that the PLL has achieved frequency lock.
After lock is detected, the lock circuit continues to monitor the reference and feedback frequencies using the alternate count and compare process. If the counters do not match at any comparison time, then the
LOCK flag is cleared to indicate that the PLL has lost lock. At this point, the lock criteria is tightened and the lock detect process is repeated.
The alternate count sequences prevent false lock detects due to frequency aliasing while the PLL tries to lock. Alternating between tight and relaxed lock criteria prevents the lock detect function from randomly toggling between locked and non-locked status due to phase sensitivities.
for detecting locked and non-locked conditions.
In external clock mode, the PLL is disabled and cannot lock.
Start with Tight Lock
Criteria
Reference Count
≠ Feedback Count
Loss of Lock Detected
Set Tight Lock Criteria and Notify System of Loss of Lock Condition
Reference Count
≠ Feedback Count
Count N
Reference Cycles and Compare
Number of Feedback
Cycles Elapsed
Reference Count =
Feedback Count = N
In Same Count/Compare Sequence
Count N + K
Reference Cycles and Compare Number of Feedback Cycles
Elapsed
Lock Detected.
Set Relaxed Lock
Condition and Notify
System of Lock
Condition
Figure 7-9. Lock Detect Sequence
Reference Count = Feedback Count = N + K
In Same Count/Compare Sequence
7-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Clock Module
7.8.3.6
PLL Loss of Lock Conditions
Once the PLL acquires lock after reset, the LOCK and LOCKS flags are set. If the MFD is changed, or if an unexpected loss of lock condition occurs, the LOCK and LOCKS flags are negated.
While the PLL is in the non-locked condition, the system clocks continue to be sourced from the PLL as the PLL attempts to relock.
Consequently, during the relocking process, the system clocks frequency is not well defined and may exceed the maximum system frequency, violating the system clock timing specifications.
However, once the PLL has relocked, the LOCK flag is set. The LOCKS flag remains cleared if the loss of lock was unexpected. The LOCKS flag is set when the loss of lock is caused by changing MFD. If the
PLL is intentionally disabled during stop mode, then after exit from stop mode, the LOCKS flag reflects the value prior to entering stop mode once lock is regained.
7.8.3.7
PLL Loss of Lock Reset
If the LOLRE bit in the SYNCR is set, a loss of lock condition asserts reset. Reset reinitializes the LOCK and LOCKS flags. Therefore, software must read the LOL bit in the reset status register (RSR) to determine if a loss of lock caused the reset. See Section 9.5.2, “Reset Status Register (RSR).”
To exit reset in PLL mode, the reference must be present, and the PLL must achieve lock.
In external clock mode, the PLL cannot lock. Therefore, a loss of lock condition cannot occur, and the
LOLRE bit has no effect.
7.8.3.8
Loss of Clock Detection
The LOCEN bit in the SYNCR enables the loss of clock detection circuit to monitor the input clocks to the phase and frequency detector (PFD). When either the reference or feedback clock frequency falls below the minimum frequency, the loss of clock circuit sets the sticky LOCS flag in the SYNSR.
NOTE
In external clock mode, the loss of clock circuit is disabled.
7.8.3.9
Loss of Clock Reset
The clock module can assert a reset when a loss of clock or loss of lock occurs. When a loss-of-clock condition is recognized, reset is asserted if the LOCRE bit in SYNCR is set. The LOCS bit in SYNSR is cleared after reset. Therefore, the LOC bit must be read in RSR to determine that a loss of clock condition occurred. LOCRE has no effect in external clock mode.
To exit reset in PLL mode, the reference must be present, and the PLL must acquire lock.
Reset initializes the clock module registers to a known startup state as described in
7.8.3.10
Alternate Clock Selection
Depending on which clock source fails, the loss-of-clock circuit switches the system clocks source to the remaining operational clock. The alternate clock source generates the system clocks until reset is asserted.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
7-15
Clock Module
The PLL remains in SCM until the next reset. When the PLL is operating in SCM, the system frequency depends on the value in the RFD field. The SCM system frequency stated in electrical specifications assumes that the RFD has been programmed to binary 000.
If the loss-of-clock condition is due to PLL failure, the PLL reference becomes the system clocks source until the next reset, even if the PLL regains and relocks.
Table 7-7. Loss of Clock Summary
Clock
Mode
PLL
System Clock Source
Before Failure
Reference Failure Alternate Clock
Selected by LOC Circuit
1
Until Reset
PLL Failure Alternate Clock
Selected by LOC Circuit Until Reset
PLL PLL self-clocked mode
None
1
The LOC circuit monitors the reference and feedback inputs to the PFD. See
.
PLL reference
NA
A special loss-of-clock condition occurs when both the reference and the PLL fail. The failures may be simultaneous, or the PLL may fail first. In either case, the reference clock failure takes priority and the
PLL attempts to operate in SCM. If successful, the PLL remains in SCM until the next reset. If the PLL cannot operate in SCM, the system remains static until the next reset. Both the reference and the PLL must be functioning properly to exit reset.
7.8.3.11
Loss of Clock in Stop Mode
shows the resulting actions for a loss of clock in stop mode when the device is being clocked by the various clocking methods.
Table 7-8. Stop Mode Operation
MODE
In
Expected
PLL
Action at
Stop
PLL Action
During Stop
EXT X X X X X X —
NRM 0 0 0 Off Off 0 Lose lock, f.b. clock, reference clock
—
Lose reference clock
Regain
No regain
MODE
Out
EXT
Stuck
NRM
Stuck
0
—
‘LK
—
0
—
0
—
1
—
‘LC
—
Comments
7-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Clock Module
Table 7-8. Stop Mode Operation (continued)
MODE
In
Expected
PLL
Action at
Stop
PLL Action
During Stop
MODE
Out
NRM X 0 0 Off Off 1 Lose lock, f.b. clock, reference clock
Regain clocks, but don’t regain lock
SCM–> unstable
NRM
No reference clock regain
SCM–>
Stuck
Comments
0–>‘LK 0–>1 1–>‘LC Block LOCS and
LOCKS until clock and lock respectively regain; enter
SCM regardless of LOCEN bit until reference regained
0–> 0–> 1–> Block LOCS and
LOCKS until clock and lock respectively regain; enter
SCM regardless of LOCEN bit
— — — No f.b. clock regain
NRM 0 0 0 Off On 0 Lose lock Regain NRM ‘LK 1 ‘LC Block LOCKS from being cleared
Lose reference clock or no lock regain
Lose reference clock, regain
NRM 0 0 0 Off On 1 Lose lock No lock regain
Stuck
NRM
Unstable
NRM
‘LK
—
1
—
‘LC
0–>‘LK 0–>1 ‘LC
—
Block LOCKS from being cleared
Block LOCKS until lock regained
Lose reference clock or no f.b. clock regain
Lose reference clock, regain
Stuck
Unstable
NRM
— —
0–>‘LK 0–>1 ‘LC
—
LOCS not set because
LOCEN = 0
NRM 0 0 0 On On 0 — — NRM
Lose lock or clock Stuck
Lose lock, regain NRM
Lose clock and lock, regain
NRM
0
0
‘LK
—
1
1
1
—
‘LC
—
‘LC
‘LC LOCS not set because
LOCEN = 0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
7-17
Clock Module
Table 7-8. Stop Mode Operation (continued)
MODE
In
Expected
PLL
Action at
Stop
PLL Action
During Stop
MODE
Out
NRM 0 0 0 On On 1 — —
Lose lock
NRM
Unstable
NRM
Lose lock, regain NRM
Lose clock Stuck
Lose clock, regain without lock
Unstable
NRM
Lose clock, regain with lock
NRM
RESET RESET NRM X X 1 Off X X Lose lock, f.b. clock, reference clock
NRM 0 0 1 On On X — — NRM
Lose lock or clock RESET
‘LK
0
0
0
0
—
—
‘LK
—
1 ‘LC
0–>1 ‘LC
1
—
‘LC
—
0–>1 ‘LC
1
—
‘LC
—
1
—
‘LC
—
NRM 1 0 0 Off Off 0 Lose lock, f.b. clock, reference clock
Regain
No regain
NRM 1 0 0 Off On 0 Lose lock, f.b. clock
Regain
NRM
Stuck
NRM
‘LK
‘LK
—
—
Comments
Reset immediately
1 ‘LC
1
—
‘LC
—
REF mode not entered during stop
— —
Reset immediately
REF not entered during stop;
SCM entered during stop only during oscillator startup
NRM 1 0 0 Off On 1 Lose lock, f.b. clock
No f.b. clock or lock regain
Lose reference clock
Stuck
SCM
Regain f.b. clock Unstable
NRM
0 0 1
0–>‘LK 0–>1 ‘LC
Wakeup without lock
REF mode not entered during stop
No f.b. clock regain
Lose reference clock
Stuck
SCM 0
—
0
—
1
—
Wakeup without lock
7-18
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Clock Module
MODE
In
NRM 1 0 0 On On 0
NRM 1 0 0 On On 1
Table 7-8. Stop Mode Operation (continued)
Expected
PLL
Action at
Stop
—
PLL Action
During Stop
—
Lose reference clock
Lose f.b. clock
MODE
Out
NRM
SCM
REF
‘LK
0
0
—
Lose lock Stuck
Lose lock, regain NRM clock
—
Lose reference
Lose f.b. clock
NRM
SCM
REF
1
0
X
‘LC
1
1
Comments
0
‘LK
0
—
0
1
1
0
— —
‘LC
‘LC
1
X 1
Wakeup without lock
Wakeup without lock
0 0–>1 ‘LC
Wakeup without lock
Wakeup without lock
NRM 1 0 1 On On X —
Lose lock Unstable
NRM
— NRM
Lose lock or clock RESET
RESET
‘LK
—
—
1
—
—
‘LC
—
—
Reset immediately
Reset immediately
NRM 1 1 X Off X X Lose lock, f.b. clock, reference clock
RESET
NRM 1 1 0 On On 0 — —
Lose clock
NRM
RESET
NRM 1 1 0 On On 1 —
Lose lock Stuck
Lose lock, regain NRM
—
Lose clock
NRM
RESET
‘LK
—
0
‘LK
—
—
0
1
—
‘LC
— Reset immediately
1
1
— —
‘LC
—
‘LC
0–>1 ‘LC
— Reset immediately
NRM 1 1 1 On On X —
Lose lock Unstable
NRM
Lose lock, regain NRM
— NRM
Lose clock or lock RESET
0
‘LK
—
1
1
—
‘LC
‘LC
— Reset immediately
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
7-19
Clock Module
Table 7-8. Stop Mode Operation (continued)
MODE
In
Expected
PLL
Action at
Stop
PLL Action
During Stop
MODE
Out
Comments
REF 1 0 0 X X X — —
Lose reference clock
Regain SCM
REF
Stuck
0
—
X
—
1
—
SCM 1 0 0 Off X 0 PLL disabled
SCM 1 0 0 Off X 1 PLL disabled
SCM 1 0 0 On On 0 —
Regain SCM
SCM
SCM
0
0
0
0
1
1
Wakeup without lock
—
Lose reference clock
—
SCM
SCM
0 0 1 Wakeup without lock
SCM 1 0 0 On On 1 —
Lose reference clock
SCM
SCM
0 0 1
Note:
PLL = PLL enabled during STOP mode. PLL = On when STPMD[1:0] = 00 or 01
OSC = oscillator enabled during STOP mode. Oscillator is on when STPMD[1:0] = 00, 01, or 10
MODES
NRM = normal PLL crystal clock reference or normal PLL external reference or PLL 1:1 mode. During PLL 1:1 or normal external reference mode, the oscillator is never enabled. Therefore, during these modes, refer to the OSC = On case regardless of STPMD values.
EXT = external clock mode
REF = PLL reference mode due to losing PLL clock or lock from NRM mode
SCM = PLL self-clocked mode due to losing reference clock from NRM mode
RESET = immediate reset
LOCKS
‘LK -= expecting previous value of LOCKS before entering stop
0–>‘LK = current value is 0 until lock is regained which then will be the previous value before entering stop
0–> = current value is 0 until lock is regained but lock is never expected to regain
LOCS
‘LC = expecting previous value of LOCS before entering stop
1–>‘LC = current value is 1 until clock is regained which then will be the previous value before entering stop
1–> = current value is 1 until clock is regained but CLK is never expected to regain
7-20
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 8
Real Time Clock
8.1
Introduction
This section discusses how to operate and program the real-time clock (RTC) module that maintains the system clock, provides stopwatch, alarm, and interrupt functions, and supports the following features.
8.1.1
Overview
is a block diagram of the functional organization of the Real Time Clock (RTC) block, it consists of the following blocks:
• Time-of-day (TOD) clock counter
• Alarm
• Minute stopwatch
• Associated control and bus interface hardware
I Hz Input CLock
RTC_INT
INTERRUPT
CONTROL
CLOCK
CONTROL
INTERRUPT
ENABLE
INTERRUPT
STATUS
ADDRESS
DATA
BUS CONTROL
IPBUS
DECODE
TOD CLOCK
SECOND
1 PPM
MINUTE
1 PPH
HOUR
1 PPD
DAY
ALARM COMPARATOR
SECOND
LATCH
MINUTE
LATCH
MINUTE STOPWATCH
HOUR
LATCH
Figure 8-1. Real Time Clock Block Diagram
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 8-1
Real Time Clock
8.1.2
Features
The RTC module includes the following features:
• Full clock—days, hours, minutes, seconds
• Minute countdown timer with interrupt
• Programmable daily alarm with interrupt
• Once-per-day, once-per-hour, once-per-minute, and once-per-second interrupts
8.1.3
Modes of Operation
The incoming 1 Hz signal is used to increment the seconds, minutes, hours, and days TOD counters. The alarm functions, when enabled, generate RTC interrupts when the TOD settings reach programmed values.
The sampling timer generates fixed-frequency interrupts, and the minute stopwatch allows for efficient interrupts on minute boundaries.
• Counter
The counter portion of the RTC module consists of four groups of counters that are physically located in three registers:
– The 6-bit seconds counter is located in the SECONDS register
– The 6-bit minutes counter and the 5-bit hours counter are located in the HOURMIN register
– The 16-bit day counter is located in the DAYR register
• Alarm
There are three alarm registers that mirror the three counter registers. An alarm is set by accessing the real-time clock alarm registers (ALRM_HM, ALRM_SEC, and DAYALARM) and loading the exact time that the alarm should generate an interrupt. When the TOD clock value and the alarm value coincide, an interrupt occurs.
• Minute Stopwatch
The minute stopwatch performs a countdown with a one minute resolution. It can be used to generate an interrupt on a minute boundary.
8.2
Memory Map/Register Definition
The RTC module includes ten 32-bit registers.
summarizes these registers and their addresses
Address
$BASE+0x00
$BASE+0x04
$BASE+0x08
$BASE+0x0C
Table 8-1. RTC Module Register Memory Map
Use
RTC Hours and Minutes Counter Register(HOURMIN)
RTC Seconds Counter Register(SECONDS)
RTC Hours and Minutes Alarm Register(ALRM_HM)
RTC Seconds Alarm Register(ALRM_SEC)
Access read/write read/write read/write read/write
8-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Memory Map/Register Definition
$BASE+0x10
$BASE+0x14
$BASE+0x18
$BASE+0x1C
$BASE+0x20
$BASE+0x24
Table 8-1. RTC Module Register Memory Map (continued)
RTC Control Register(RCCTL)
RTC Interrupt Status Register(RTCISR)
RTC Interrupt Enable Register(RTCIENR)
Stopwatch Minutes Register(STPWCH)
RTC Days Counter Register(DAYS)
RTC Seconds Counter Register(ALARM_DAY) read/write read/write read/write read/write read/write read/write
8.2.1
Register Descriptions
This section consists of register descriptions in address order. Each description includes a standard register diagram with an associated figure number. Details of register bit and field function follow the register diagrams, in bit order.
8.2.1.1
RTC Hours and Minutes Counter Register(HOURMIN)
The real-time clock hours and minutes counter register (HOURMIN) is used to program the hours and minutes for the TOD clock. It can be read or written at any time. After a write, the time changes to the new value. This register cannot be reset since the real-time clock is always enabled at reset.
$BASE_ADDRESS+0x0
31
0
30
0 R
W
RESET:
29
0
28
0
27
0
26
0
25
0
24
0
23
0
0x0000
22
0
21
0
20
0
19
0
18
0
17
0
16
0
R
W
RESET:
15
0
0
14
0
0
13
0
0
12
?
11 10
HOURS
?
?
9
?
8
?
7
0
0
6
0
0
5
?
4
?
3 2
MINUTES
?
?
1
?
= Unimplemented or Reserved
Figure 8-2. RTC Hours and Minutes Counter Register
HOURS — Hour setting which indicates the current hour, can be set to any value between 0 and 23.
Table 8-2. Meanings of HOURS (HOURMIN Register)
Values
00000
00001
......
10111
Meanings current hour is 0 current hour is 1
......
current hour is 23
0
?
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 8-3
Real Time Clock
MINUTES — Minutes setting which indicates the current minutes, can be set to any value between 0 and
59.
Table 8-3. Meanings of MINUTES(HOURMIN Register)
Values
000000
000001
......
111011
Meanings current minute is 0 current minute is 1
......
current minute is 59
8.2.1.2
RTC Seconds Counter Register(SECONDS)
The real-time clock seconds register (SECONDS) is used to program the seconds for the TOD clock. It can be read or written at any time. After a write, the time changes to the new value. This register cannot be reset since the real-time clock is always enabled at reset.
$BASE_ADDRESS+0x04
31
0
30
0 R
W
RESET:
29
0
28
0
27
0
26
0
25
0
24
0
0x0000
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
R
W
RESET:
15
0
0
14
0
0
13
0
0
12
0
0
11
0
0
10
0
0
9
0
0
8
0
0
7
0
0
6
0
0
5
?
4
?
3 2
SECONDS
?
?
1
?
0
?
= Unimplemented or Reserved
Figure 8-3. RTC Seconds Counter Register
SECONDS—Seconds setting which indicates the current seconds, can be set to any value between 0 and
59.
Table 8-4. Meanings of SECONDS(SECONDS Register)
Values
000000
000001
......
111011
Meanings current second is 0 current second is 1
......
current second is 59
8.2.1.3
RTC Hours and Minutes Alarm Register(ALRM_HM)
The real-time clock hours and minutes alarm (ALRM_HM) register is used to configure the hours and minutes setting for the alarm. The alarm settings can be read or written at any time.
8-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Memory Map/Register Definition
$BASE_ADDRESS+0x08
31
0
30
0
29
0 R
W
RESET:
28
0
27
0
26
0
25
0
24
0
0x0000
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
R
W
RESET:
15
0
14
0
13
0
12 11 10
HOURS
9 8
0x0000
7
0
6
0
5 4 3 2
MINUTES
= Unimplemented or Reserved
Figure 8-4. RTC Hours and Minutes Alarm Register
HOURS — Hour setting of the alarm hours, can be set to any value between 0 and 23.
Table 8-5. Meanings of HOURS(ALRM_HM Register)
Values
00000
00001
......
10111
Meanings current hour is 0 current hour is 1
......
current hour is 23
MINUTES — Minutes setting of the alarm minutes, can be set to any value between 0 and 59.
Table 8-6. Meanings of MINUTES(ALRM_HM Register)
Values
000000
000001
......
111011
Meanings current minute is 0 current minute is 1
......
current minute is 59
1 0
8.2.1.4
RTC Seconds Alarm Register(ALRM_SEC)
The real-time clock seconds alarm (ALRM_SEC) register is used to configure the seconds setting for the alarm. The alarm settings can be read or written at any time.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 8-5
Real Time Clock
$BASE_ADDRESS+0x0C
31
0
30
0
29
0 R
W
RESET:
28
0
27
0
26
0
25
0
24
0
0x0000
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
R
W
RESET:
15
0
14
0
13
0
12
0
11
0
10
0
9
0
8
0
0x0000
7
0
6
0
5 4 3 2
SECONDS
= Unimplemented or Reserved
Figure 8-5. RTC Seconds Alarm Register
SECONDS—Seconds setting of the alarm seconds, can be set to any value between 0 and 59.
Table 8-7. Meanings of SECONDS(ALRM_SEC Register)
Values
000000
000001
......
111011
Meanings current second is 0 current second is 1
......
current second is 59
1 0
8.2.1.5
RTC Control Register(RTCCTL)
The real-time clock control (RTCCTL) register is used to enable the real-time clock module and specify the reference frequency information for the prescaler.
$BASE_ADDRESS+0x10
31
0
30
0 R
W
RESET:
29
0
28
0
27
0
26
0
25
0
24
0
0x0000
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
R
W
RESET:
15
0
14
0
13
0
12
0
11
0
10
0
9
0
8
0
7
EN
0x0080
6
0
= Unimplemented or Reserved
Figure 8-6. RTC Control Register
5
0
4
0
3
0
2
0
1
0
0
SWR
8-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Memory Map/Register Definition
EN — Enables/Disables the real-time clock. The software reset bit (SWR) has no effect on this bit.
Bit description
1 = Enable the real-time clock
0 = Disable the real-time clock
XTL — Crystal Selection, selects the proper input crystal frequency. It is important to set these bits correctly or the real-time clock will be inaccurate.
Table 8-8. Meanings of XTL(RTCCTL Register)
Values
00
01
10
11
Meanings input crystal frequency is 32.768KHz
input crystal frequency is 32KHz input crystal frequency is 38.4KHz
input crystal frequency is 32.768KHz
SWR — Software reset, Resets the module to its default state. However, a software reset will have no effect on the RTC enable (EN) bit.
Bit description
1 = Reset the module to its default state.
0 = No effect.
8.2.1.6
RTC Interrupt Status Register(RTCISR)
The real-time clock interrupt status register (RTCISR) indicates the status of the various real-time clock interrupts. When an event of the types included in this register occurs, then the bit will be set in this register regardless of its corresponding interrupt enable bit.These bits are cleared by writing a value of 1, which also clears the interrupt. Interrupts may occur while the system clock is idle or in sleep mode.
$BASE_ADDRESS+0x14
31
0
30
0 R
W
RESET:
29
0
28
0
27
0
26
0
25
0
24
0
0x0000
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
R
W
RESET:
15
0
14
0
13
0
12
0
11
0
10
0
9
0
8
0
0x0000
7
0
6
0
5 4 3 2 1 0
HR 1HZ DAY ALM MIN SW
= Unimplemented or Reserved
Figure 8-7. RTC Interrupt Status Register
HR — Hour Flag, indicates that the hour counter has incremented. If enabled, this bit is set on every increment of the hour counter in the time-of-day clock.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 8-7
Real Time Clock
Bit description
1 = A 1-hour interrupt has occurred.
0 = No 1-hour interrupt occurred.
1HZ — 1 Hz Flag, indicates that the second counter has incremented. If enabled, this bit is set on every increment of the second counter of the time-of-day clock.
Bit description
1 = A 1 Hz interrupt has occurred.
0 = No 1 Hz interrupt occurred.
DAY — Day Flag, indicates that the day counter has incremented. If enabled, this bit is set on every increment of the day counter of the time-of-day clock.
Bit description
1 = A 24-hour rollover interrupt has occurred.
0 = No 24-hour rollover interrupt occurred.
ALM — Alarm Flag—Indicates that the real-time clock matches the value in the alarm registers. Note that the alarm will reoccur every 65536 days. For a single alarm, clear the interrupt enable for this bit in the interrupt service routine.
Bit description
1 = An alarm interrupt has occurred.
0 = No alarm interrupt occurred.
MIN — Minute Flag, indicates that the minute counter has incremented. If enabled, this bit is set on every increment of the minute counter in the time-of-day clock.
Bit description
1 = A 1-minute interrupt has occurred.
0 = No 1-minute interrupt occurred.
SW — Stopwatch Flag, indicates that the stopwatch countdown timed out.
Bit description
1 = The stopwatch timed out.
0 = The stopwatch did not time out.
8.2.1.7
RTC Interrupt Enable Register(RTCIENR)
The real-time clock interrupt enable register (RTCIENR) is used to enable/disable the various real-time clock interrupts. Masking an interrupt bit has no effect on its corresponding status bit.
8-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Memory Map/Register Definition
$BASE_ADDRESS+0x18
31
0
30
0 R
W
RESET:
29
0
28
0
27
0
26
0
25
0
24
0
0x0000
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
R
W
RESET:
15
0
14
0
13
0
12
0
11
0
10
0
9
0
8
0
0x0000
7
0
6
0
5 4 3 2 1 0
HR 1HZ DAY ALM MIN SW
= Unimplemented or Reserved
Figure 8-8. RTC Interrupt Enable Register
HR — Hour Interrupt Enable, Enables/Disables an interrupt whenever the hour counter of the real-time clock increments.
Bit description
1 = The 1-hour interrupt is enabled.
0 = The 1-hour interrupt id disabled.
1HZ — 1 Hz Interrupt Enable, Enables/Disables an interrupt whenever the second counter of the real-time clock increments.
Bit description
1 = The 1 Hz interrupt is enabled.
0 = The 1 Hz interrupt is disabled.
DAY — Day Interrupt Enable, Enables/Disables an interrupt whenever the hours counter rolls over from
23 to 0. (midnight rollover)
Bit description
1 = The 24-hour interrupt is enabled.
0 = The 24-hour interrupt is disabled.
ALM — Alarm Interrupt Enable, Enables/Disables the alarm interrupt.
Bit description
1 = The alarm interrupt is enabled.
0 = The alarm interrupt is disabled.
MIN — Minute Interrupt Enable, Enables/Disables an interrupt whenever the minute counter of the real-time clock increments.
Bit description
1 = The 1-minute interrupt is enabled.
0 = The 1-minute interrupt is disabled.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 8-9
Real Time Clock
SW — Stopwatch Interrupt Enable, Enables/Disables the stopwatch interrupt. Please note that the stopwatch counts down and remains at decimal -1 until it is reprogrammed. If this bit is enabled with -1
(decimal) in the STPWCH register, an interrupt will be posted on the next minute tick.
Bit description
1 = Stopwatch interrupt is enabled.
0 = Stopwatch interrupt is disabled.
8.2.1.8
RTC Stopwatch Minutes Register(STPWCH)
The stopwatch minutes (STPWCH) register contains the current stopwatch countdown value. When the minute counter of the TOD clock increments, the value in this register decrements.
$BASE_ADDRESS+0x1C
31
0
30
0
29
0 R
W
RESET:
28
0
27
0
26
0
25
0
24
0
0x0000
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
R
W
RESET:
15
0
14
0
13
0
12
0
11
0
10
0
9
0
8
0
0x003F
7
0
6
0
5 4 3
CNT
2 1 0
= Unimplemented or Reserved
Figure 8-9. RTC Stopwatch Minutes Register
CNT — Stopwatch Count, Contains the stopwatch countdown value. Please note that the stopwatch counter is decremented by the minute (MIN) tick output from the real-time clock, so the average tolerance of the count is 0.5 minutes. For better accuracy, enable the stopwatch by polling the MIN bit of the
RTCISR register or by polling the minute interrupt service routine.
Table 8-9. Meanings of CNT(STPWCH Register)
Values
000000
000001
......
111111
Meanings stopwatch countdown value is 0 stopwatch countdown value is 1
......
stopwatch countdown value is 63
8.2.1.9
RTC Days Counter Register(DAYR)
The real-time clock days counter register (DAYR) is used to program the day for the TOD clock. When the HOUR field of the HOURMIN register rolls over from 23 to 00, the day counter increments. It can be read or written at any time. After a write, the time changes to the new value. This register cannot be reset since the real-time clock is always enabled at reset. Only 16-bit accesses to this register are allowed.
8-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Memory Map/Register Definition
$BASE_ADDRESS+0x20
31
0
30
0 R
W
RESET:
29
0
28
0
27
0
26
0
25
0
24
0
0x0000
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
R
W
RESET:
15 14 13 12 11 10 9 8
DAYS
7
0x????
6 5 4 3 2 1
= Unimplemented or Reserved
Figure 8-10. RTC Days Counter Register
DAYS — Day Setting, indicates the current day count, can be set to any values between 0 and 65535.
Table 8-10. Meanings of DAYS(DAYR Register)
Values
0x0000
0x0001
......
0xFFFF
Meanings current day count is 0 current day count is 1
......
current day count is 65535
0
8.2.1.10
RTC Day Alarm Register(DAYALARM)
The real-time clock day alarm (DAYALARM) register is used to configure the day for the alarm. The alarm settings can be read or written at any time.
$BASE_ADDRESS+0x24
31
0
30
0 R
W
RESET:
29
0
28
0
27
0
26
0
25
0
24
0
0x0000
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
R
W
RESET:
15 14 13 12 11 10 9 8 7
DAYSAL
0x0000
6 5 4 3 2 1 0
= Unimplemented or Reserved
Figure 8-11. RTC Day Alarm Register
DAYSAL — Day Setting of the Alarm, indicates the current day setting of the alarm. It can be set to any value between 0 and 65535.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 8-11
Real Time Clock
Table 8-11. Meanings of DAYSAL(DAYALARM Register)
Values
0x0000
0x0001
......
0xFFFF
Meanings current day setting of alarm is 0 current day setting of alarm is 1
......
current day setting of alarm is 65535
8.3
Functional Description
A 1 Hz signal is supplied to the RTC, which used it to increment the seconds, minutes, hours, and days
TOD counters. The alarm functions, when enabled, generate RTC interrupts when the TOD settings reach programmed values. The minute stopwatch allows for efficient interrupts on minute boundaries.
8.3.1
Prescaler and Counter
A 1 Hz clock is supplied to the RTC. This 1 Hz clock drives the RTC’s counters. The counter portion of the RTC module consists of four groups of counters that are physically located in three registers:
• The 6-bit seconds counter is located in the SECONDS register
• The 6-bit minutes counter and the 5-bit hours counter are located in the HOURMIN register
• The 16-bit day counter is located in the DAYR register
These counters cover a 24-hour clock over 65536 days. All three registers can be read or written at any time.
Interrupts signal when each of the four counters increments, and can be used to indicate when a counter rolls over. For example, each tick of the seconds counter causes the 1HZ interrupt flag to be set. When the seconds counter rolls from 59 to 00, the minute counter increments and the MIN interrupt flag is set. The same is true for the minute counter with the HR signal, and the hour counter with the DAY signal.
8.3.2
Alarm
There are three alarm registers that mirror the three counter registers. An alarm is set by accessing the real-time clock alarm registers (ALRM_HM, ALRM_SEC, and DAYALARM) and loading the exact time that the alarm should generate an interrupt. When the TOD clock value and the alarm value coincide, if the ALM bit in the real-time clock interrupt enable register (RTCIENR) is set, an interrupt occurs. Please be noted that if the alarm is not disabled, it will reoccur every 65536 days. If a single alarm is desired, the alarm function must be disabled through the RTC Interrupt Enable Register (RTCIENR).
8.3.3
Minute Stopwatch
The minute stopwatch performs a countdown with a one minute resolution. It can be used to generate an interrupt on a minute boundary. For example, to turn off the LCD controller after five minutes of inactivity, program a value of 0x04 into the Stopwatch Count (CNT) field of the Stopwatch Minutes (STPWCH)
8-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Initialization/Application Information register. At each minute, the value in the stopwatch is decremented. When the stopwatch value reaches -1, the interrupt occurs. The value of the register does not change until it is reprogrammed. Note that the actual delay includes the seconds from setting the stopwatch to the next minute tick.
8.4
Initialization/Application Information
8.4.1
Flow Chart of RTC Operation
shows the flow chart of a typical RTC operation.
configure RTC Control Register c onfig RTC Days Counter Register config RTC Seconds Counter Reg config RTC Hr/Min Counter Register config RTC Alarm Registers config RTC Interrupt Enable Reg
Check RTC Interrupt Status Register
Figure 8-12. Flow Chart of RTC Operation
8.4.2
Code Example to initialize the Real-Time Clock
is the code example of instructions for configuring RTC.
LDR r1,=RTC_BASE_ADDR
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 8-13
Real Time Clock
LDR r2,[r1,#0x10]
ORR r2,r2,#0x21
STR r2,[r1,#0x10] ; Software reset and 32k crystal
LDR r3,=0x0000
STR r3,[r1,#0x20] ;DAY
LDR r3,=0x00038
STR r3,[r1,#0x04] ;SECOND
LDR r3,=0x173B
STR r3,[r1] ;HR, MIN
LDR r3,=0x0001
STR r3,[r1,#0x24] ;Alarm Day
LDR r3,=0x0000
STR r3,[r1,#0x08] ;Alarm hour, minute
LDR r3,=0x01
STR r3,[r1,#0x0C] ;Alarm seconds
LDR r2,[r1,#0x18] ;set ALARM interrupt
ORR r2,r2,#0x4
STR r2,[r1,#0x18]
ALARM_STATUS_3
LDR r2,[r1,#0x18] ;check ALARM STATUS FLAG
TST r2,#0x04
BNE ALARM_STATUS_3
Figure 8-13. Code Example of ARM Instruction
8-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Chapter 9
Power Management
9.1
Introduction
This chapter explains the low-power operation of the MCF5235.
9.1.1
Features
The following features support low-power operation.
• Four modes of operation: run, wait, doze, and stop
• Ability to shut down most peripherals independently
• Ability to shut down the external CLKOUT pin
9.2
Memory Map/Register Definition
The power management programming model consists of registers from both the SCM and CCM memory
space, as shown in Table 9-1 .
Table 9-1. Power Management Memory Map
IPSBAR
Offset
1
Register
Width
(bits)
Access Reset Value Section/Page
0x11_0004 Chip Configuration Register (CCR)
2
0x11_0007 Low-Power Control Register (LPCR)
16
8
R
R/W
0x0001
0x00
0x00_000C Peripheral Power Management Register High (PPMRH)
0x00_0010 Core Reset Status Register (CRSR)
3
0x00_0011 Core Watchdog Control Register (CWCR)
3
0x00_0012 Low-Power Interrupt Control Register (LPICR)
0x00_0013 Core Watchdog Service Register (CWSR)
3
32
8
8
8
R/W
R/W
R/W
R/W
0x00000000
0x00
0x00
0x00_0018 Peripheral Power Management Register Low (PPMRL)
0x00_0021 Peripheral Power Management Set Register (PPMRS)
8
32
8
R/W
R/W
W
0x00000001
0x00
0x00_0022 Peripheral Power Management Clear Register (PPMRC) 32 R/W 0x00
1
2
Addresses not assigned to a register and undefined register bits are reserved for expansion.
The CCR is described in the Chip Configuration Module. It is shown here only to warn against accidental writes to this register when accessing the LPCR.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 9-1
Power Management
3
The CRSR, CWCR, and CWSR are described in the System Control Module. They are shown here only to warn against accidental writes to these registers when accessing the LPICR.
Table 9-2.
IPSBAR Offset [31:24] [23:16] [15:8] [7:0]
0x00_000C
0x00_0010 Core Reset Status
Register (CRSR)
1
PPMRH[63:32]
Core Watchdog Control
Register (CWCR)
3
Low-Power Interrupt
Control Register
(LPICR)
Core Watchdog
Service Register
(CWSR)
3
0x00_0018 PPMRL[31:0]
0x00_0020
0x11_0004
— PPMRS
Chip Configuration Register (CCR)
2
PPMRC
Reserved
—
Low-Power Control
Register (LPCR)
1
2
The CCR is described in the Chip Configuration Module. It is shown here only to warn against accidental writes to this register when accessing the LPCR.
9.2.1
Peripheral Power Management Registers (PPMRH, PPMRL)
The PPMRH and PPMRL registers provide a bit map for controlling the generation of the module clocks for each decoded address space associated with the IPS controller. The PPMR x provides a unique control bit for each of these address spaces that defines whether the module clock for the given space is enabled or disabled.
NOTE
It is software’s responsibility to appropriately disable module clocks using the PPMR x only when a module is completely unused or quiescent.
Since the operation of the IPS controller and the system control module (SCM) are fundamental to the operation of the system, the clocks for these three modules cannot be disabled.
The individual bits of the PPMR x can be modified using a read-modify-write to this register directly or indirectly through writes to the PPMRS and PPMRC registers to set/clear individual bits.
See Figure 7-2 and
for the PPMRH definition.
9-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Power Management
IPSBAR
Offset:
0x00_000C (PPMRH)
31
0 R
W
Reset 0
30
0
0
23
0
22
0 R
W
Reset 0
15
0
0
14
0 R
W
Reset 0 0
29
0
0
21
0
0
13
CDRNGA
0
28
0
0
20
0
0
12
CDEPHY
0
27
0
0
19
0
0
11
CDCFM
0
26
0
0
18
0
0
10
CDFCAN
0
25
0
0
17
0
Access: read/write
0
9
CDPWM
0
24
0
0
16
0
0
8
CDGPT
0
R
W
Reset
7
CDADC
0
6
0
5
0
4
CDPIT1
3
CDPIT0
2
0
0 0 0 0 0
Figure 9-1. Peripheral Power Management Register High (PPMRH)
1 0
CDEPORT CDPORTS
0 0
Field
31–14
13
CDEPHY
12
CDRNGA
11
CDCFM
10
CDFCAN
9
CDPWM
8
CDGPT
Table 9-3. PPMRH Field Descriptions
Description
Reserved, should be cleared.
Disable clock to the EPHY (Ethernet PHY Module))
0 EPHY module clock is enabled
1 EPHY module clock is disabled
Disable clock to the RNGA (Random Number Generator Accelerator Module)
0 RNGA module clock is enabled
1 RNGA module clock is disabled
Disable clock to the CFM (Common Flash Module)
0 CFM module clock is enabled
1 CFM module clock is disabled
Disable clock to the FlexCAN module.
0 FlexCAN module clock is enabled
1 FlexCAN module clock is disabled
Disable clock to the PWM module.
0 PWM module clock is enabled
1 PWM module clock is disabled
Disable clock to the 16 bit general purpose timer module (GPT).
0 ICOC module clock is enabled
1 ICOC module clock is disabled
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 9-3
Power Management
Field
7
CDADC
6–5
4
CDPIT1
3
CDPIT0
2
1
CDEPORT
0
CDPORTS
Table 9-3. PPMRH Field Descriptions (continued)
Description
Disable clock to the ADC module.
0 ADC module clock is enabled
1 ADC module clock is disabled
Reserved, should be cleared.
Disable clock to the PIT1 module.
0 PIT0 module clock is enabled
1 PIT1 module clock is disabled
Disable clock to the PIT0 module.
0 PIT0 module clock is enabled
1 PIT0 module clock is disabled
Reserved, should be cleared.
Disable clock to both EPORT modules.
0 EPORT module clock is enabled
1 EPORT module clock is disabled
Disable clock to the Ports module.
0 Ports module clock is enabled
1 Ports module clock is disabled
9.2.1.1
Peripheral Power Management Register Low (PPMRL)
IPSBAR
Offset:
0x00_0018 (PPMRL)
31
0 R
W
Reset 0
30
0
0
29
0
0
28
0
0
27
0
0
26
0
0
25
0
Access: read/write
0
R
W
Reset
23
0
0
R
W
Reset
15
CDTMR2
0
22
0
0
14
CDTMR1
0
21
CDFEC0
0
13
CDTMR0
0
20
0
0
12
CDRTC
0
19
0
0
11
0
18
CDINTC1
0
10
CDQSPI
0
17
CDINTC0
0
9
CDI2C
0 0
R
W
Reset
7
CDUART2
0
6
CDUART1
5
CDUART0
4
CDDMA
3
0
2
0
0 0 0 1 0
Figure 9-2. Peripheral Power Management Register Low (PPMRL)
1
CDG
0
24
0
0
16
CDTMR3
0
8
0
0
0
0
0
9-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Field
31–22
21
CDFEC0
20–19
18
CDINTC0
17
CDINTC0
16
CDTMR3
15
CDTMR2
14
CDTMR1
13
CDTMR0
12
CDRTC
11
10
CDQSPI
9
CDI2C
8
—
7
CDUART2
6
CDUART1
Table 9-4. PPMRL Field Descriptions
Description
Reserved, should be cleared.
Disable clock to the FEC (Fast Ethernet Controller) module.
0 FEC module clock is enabled
1 FEC module clock is disabled
Reserved, should be cleared.
Disable clock to the INTC1 module.
0 INTC1 module clock is enabled
1 INTC1 module clock is disabled
Disable clock to the INTC0 module.
0 INTC0 module clock is enabled
1 INTC0 module clock is disabled
Disable clock to the DTIM3 module.
0 TMR3 module clock is enabled
1 TMR3 module clock is disabled
Disable clock to the DTIM2 module.
0 TMR2 module clock is enabled
1 TMR2 module clock is disabled
Disable clock to the DTIM1 module.
0 TMR1 module clock is enabled
1 TMR1 module clock is disabled
Disable clock to the DTIM0 module.
0 TMR0 module clock is enabled
1 TMR0 module clock is disabled
Disable clock to the RTC (Real-Time Clock) module.
0 RTC module clock is enabled
1 RTC module clock is disabled
Reserved, should be cleared.
Disable clock to the QSPI module.
0 QSPI module clock is enabled
1 QSPI module clock is disabled
Disable clock to the I2C module.
0 I2C module clock is enabled
1 I2C module clock is disabled
Reserved, should be cleared.
Disable clock to the UART2 module.
0 UART1 module clock is enabled
1 UART2 module clock is disabled
Disable clock to the UART1 module.
0 UART1 module clock is enabled
1 UART1 module clock is disabled
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary
Power Management
9-5
Power Management
Field
5
CDUART0
4
CDDMA
3–2
1
CDG
0
Table 9-4. PPMRL Field Descriptions (continued)
Description
Disable clock to the UART0 module.
0 UART0 module clock is enabled
1 UART0 module clock is disabled
Disable clock to the DMA module.
0 DMA module clock is enabled
1 DMA module clock is disabled
Reserved, should be cleared.
Disable clock to the Global off-platform modules.
0 Global off-platform module clocks are enabled
1 Global off-platform module clocks are disabled
Reserved, should be cleared.
9.2.2
Low-Power Interrupt Control Register (LPICR)
Implementation of low-power stop mode and exit from a low-power mode via an interrupt require communication between the CPU and logic associated with the interrupt controller. The LPICR is an 8-bit register that enables entry into low-power stop mode, and includes the setting of the interrupt level needed to exit a low-power mode.
NOTE
The setting of the low-power mode select (LPMD) field in the power management module’s low-power control register (LPCR) determines which low-power mode the device enters when a STOP instruction is issued.
If this field is set to enter stop mode, then the ENBSTOP bit in the LPICR must also be set.
The following is the sequence of operations needed to enable this functionality:
1. The LPICR is programmed, setting the ENBSTOP bit (if stop mode is the desired low-power mode) and loading the appropriate interrupt priority level.
2. At the appropriate time, the processor executes the privileged STOP instruction. Once the processor has stopped execution, it asserts a specific Processor Status (PST) encoding. Issuing the
STOP instruction when the LPICR[ENBSTOP] bit is set causes the SCM to enter stop mode.
3. The entry into a low-power mode is processed by the low-power mode control logic, and the appropriate clocks (usually those related to the high-speed processor core) are disabled.
4. After entering the low-power mode, the interrupt controller enables a combinational logic path which evaluates any unmasked interrupt requests. The device waits for an event to generate an interrupt request with a priority level greater than the value programmed in
LPICR[XLPM_IPL[2:0]].
9-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Power Management
NOTE
Only a fixed (external) interrupt can bring a device out of stop mode. To exit from other low-power modes, such as doze or wait, either fixed or programmable interrupts may be used; however, the module generating the interrupt must be enabled in that particular low-power mode.
5. Once an appropriately high interrupt request level arrives, the interrupt controller signals its presence, and the SCM responds by asserting the request to exit low-power mode.
6. The low-power mode control logic senses the request signal and re-enables the appropriate clocks.
7. With the processor clocks enabled, the core processes the pending interrupt request.
Access: read/write IPSBAR
Offset:
0x00_0012 (LPICR)
7
R
W
Reset:
ENBSTOP
0
6 5
XLPM_IPL[2:0]
4 3
0
2
0
0 0 0 0 0
Figure 9-3. Low-Power Interrupt Control Register (LPICR)
1
0
0
0
0
0
Table 9-5. LPICR Field Description
Field Description
7
ENBSTOP
Enable low-power stop mode.
0 Low-power stop mode disabled
1 Low-power stop mode enabled. Once the core is stopped and the signal to enter stop mode is asserted, processor clocks can be disabled.
6–4
XLPM_IPL
[2:0]
Exit low-power mode interrupt priority level. This field defines the interrupt priority level needed to exit the low-power mode.Refer to
3–0
—
Reserved, should be cleared.
XLPM_IPL[2:0]
000
001
010
011
100
101
11x
Table 9-6. XLPM_IPL Settings
Interrupts Level Needed to Exit Low-Power Mode
Any interrupt request exits low-power mode
Interrupt request levels 2–7 exit low-power mode
Interrupt request levels 3–7 exit low-power mode
Interrupt request levels 4–7 exit low-power mode
Interrupt request levels 5–7 exit low-power mode
Interrupt request levels 6–7 exit low-power mode
Interrupt request level 7 exits low-power mode
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 9-7
Power Management
9.2.3
Peripheral Power Management Set Register (PPMRS)
The PPMRS register provides a simple memory-mapped mechanism to set a given bit in the PPMR x registers to disable the clock for a given IPS module without the need to perform a read-modify-write on the PPMR. The data value on a register write causes the corresponding bit in the PPMR x register to be set.
A data value of 64 to 127 provides a global set function, forcing the entire contents of the PPMR x to be set, disabling all IPS module clocks. Reads of this register return all zeroes. See
and
for the PPMRS definition.
IPSBAR
Offset:
0x00_0021 (PPMRS)
R
W
Reset:
7
0
6 5 4 3 2 1
0 0 0 0
PPMRS
0 0
Figure 9-4. Peripheral Power Management Set Register (PPMRS)
0
Access: write-only
0
0
Field
7
6–0
PPMRS
Table 9-7. PPMRS Field Descriptions
Description
Reserved, should be cleared.
Set Module Clock Disable
0–63 Set corresponding bit in PPMR x , disabling the module clock
64–127 Set all bits in PPMR x , disabling all the module clocks
9.2.4
Peripheral Power Management Clear Register (PPMRC)
The PPMRC register provides a simple memory-mapped mechanism to clear a given bit in the PPMR x registers to enable the clock for a given IPS module without the need to perform a read-modify-write on the PPMR x . The data value on a register write causes the corresponding bit in the PPMR x register to be cleared. A data value of 64 to 127 provides a global clear function, forcing the entire contents of the
PPMR x to be zeroed, enabling all IPS module clocks. In the event on simultaneous writes of the PPMRS
and Table 9-8 for the PPMRC definition.
IPSBAR
Offset:
0x00_0022 (PPMRC)
R
W
Reset:
7
0
6 5 4 3 2 1
0 0 0 0
PPMRC
0 0
Figure 9-5. Peripheral Power Management Clear Register (PPMRC)
0
Access: write-only
0
0
9-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Power Management
Field
7
6–0
PPMRC
Table 9-8. PPMRC Field Descriptions
Description
Reserved, should be cleared.
Clear Module Clock Disable
0–63 Clear corresponding bit in PPMR x , enabling the module clock
64–127 Clear all bits in PPMR x , enabling all the module clocks
9.2.4.1
Low-Power Control Register (LPCR)
The LPCR controls chip operation and module operation during low-power modes. The low-power control register (LPCR) specifies the low-power mode entered when the STOP instruction is issued, and controls clock activity in this low-power mode.
Access: read/write IPSBAR
Offset:
0x11_0007 (LPCR)
7
R
W
Reset: 0
LPMD
6 5
0
4
0
3
STPMD
0 0 0 0 0
Figure 9-6. Low-Power Control Register (LPCR)
2
0
1
0
0
0
0
0
Field
7–6
LPMD
Table 9-9. LPCR Field Descriptions
Description
Low-power mode select. Used to select the low-power mode the chip enters once the ColdFire CPU executes the
STOP instruction. These bits must be written prior to instruction execution for them to take effect. The LPMD[1:0] bits are readable and writable in all modes. Below illustrates the four different power modes that can be configured with the LPMD bit field.
5–4
3
STPMD
2–0
LPMD[1:0]
11
10
01
00
Mode
STOP
WAIT
DOZE
RUN
Note: If LPCR[LPMD] is cleared, then the device will stop executing code upon issue of a STOP instruction.
However, no clocks will be disabled.
Reserved, should be cleared.
CLKOUT stop mode. Controls CLKOUT operation during stop mode.
0 CLKOUT enabled during stop mode.
1 CLKOUT disabled during stop mode.
Reserved, should be cleared.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 9-9
Power Management
9.3
IPS Bus Timeout Monitor
The IPS controller implements a bus timeout monitor to insure that every IPS bus cycle is properly terminated within a programmed period of time. The monitor continually checks for termination of each
IPS bus cycle and completes the cycle if there is no response when the programmed monitor cycle count is reached. The error termination is propagated onto the system bus and eventually back to the ColdFire
Core.
The monitor can be programmed from 8–1024 system bus cycles under control of the IPS Bus Monitor
Timeout Register (IPSBMT). The timeout value must be selected so that it is larger than the response time of the slowest IPS peripheral device. The bus timeout monitor begins counting on the initial assertion of any IPS module enable and continues to count until the bus cycle is terminated via the negation of ips_xfr_wait . If the programmed timeout value is reached before a termination, the bus monitor completes the cycle with an error termination. At reset, the IPSBMT is enabled with a maximum timeout value. See
Figure 9-7 and Table 9-10 for the IPSBMT definition.
0x00_0023 (IPSBMT)
31
R 0
W
Reset 0
30
0
0
29
0
0
28
0
0
27
0
0
26
0
0
25
0
0
24
0
0
23
0
0
22
0
0
21
0
0
20
0
0
19
0
0
18
0
Access: read/write
17
0
16
0
0 0 0
15
R 0
W
Reset 0
14
0
0
13
0
12
0
11
0
10
0
9
0
8
0
7
0
6
0
5
0
4
0
0 0 0 0 0 0 0 0 0 0
Figure 9-7. IPS Bus Timeout Monitor (IPSBMT) Register
3
BME
1
2
0
1
BMT
0
0
0
Field
15–4
3
BME
2–0
BMT[2:0]
Table 9-10. IPSBMT Field Description
Description
Reserved, should be cleared.
Bus Timeout Monitor Enable
0 The bus timeout monitor is disabled.
1 The bus timeout monitor is enabled.
Bus Monitor Timeout. This field selects the timeout period (measured in system bus clock cycles) for the bus monitor.
000 1024 cycles
001 512 cycles
010 256 cycles
011 128 cycles
100 64 cycles
101 32 cycles
110 16 cycles
111 8 cycles
9-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Power Management
9.4
Functional Description
The functions and characteristics of the low-power modes, and how each module is affected by, or affects these modes are discussed in this section.
9.4.1
Low-Power Modes
The system enters a low-power mode by executing a STOP instruction. Which mode the device actually enters (either stop, wait, or doze) depends on what is programmed in LPCR[LPMD]. Entry into any of these modes idles the CPU with no cycles active, powers down the system and stops all internal clocks appropriately. During stop mode, the system clock is stopped low.
For entry into stop mode, the LPICR[ENBSTOP] bit must be set before a STOP instruction is issued.
A wakeup event is required to exit a low-power mode and return to run mode. Wakeup events consist of any of these conditions:
• Any type of reset
• Any valid, enabled interrupt request
Exiting from low power mode via an interrupt request requires:
• An interrupt request whose priority is higher than the value programmed in the XLPM_IPL field of the LPICR.
• An interrupt request whose priority higher than the value programmed in the interrupt priority mask (I) field of the core’s status register.
• An interrupt request from a source which is not masked in the interrupt controller’s interrupt mask register.
• An interrupt request which has been enabled at the module of the interrupt’s origin.
9.4.1.1
Run Mode
Run mode is the normal system operating mode. Current consumption in this mode is related directly to the system clock frequency.
9.4.1.2
Wait Mode
Wait mode is intended to be used to stop only the CPU and memory clocks until a wakeup event is detected. In this mode, peripherals may be programmed to continue operating and can generate interrupts, which cause the CPU to exit from wait mode.
9.4.1.3
Doze Mode
Doze mode affects the CPU in the same manner as wait mode, except that each peripheral defines individual operational characteristics in doze mode. Peripherals which continue to run and have the capability of producing interrupts may cause the CPU to exit the doze mode and return to run mode.
Peripherals which are stopped will restart operation on exit from doze mode as defined for each peripheral.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 9-11
Power Management
9.4.1.4
Stop Mode
Stop mode affects the CPU in the same manner as the wait and doze modes, except that all clocks to the system are stopped and the peripherals cease operation.
Stop mode must be entered in a controlled manner to ensure that any current operation is properly terminated. When exiting stop mode, most peripherals retain their pre-stop status and resume operation.
The following subsections specify the operation of each module while in and when exiting low-power modes.
9.4.1.5
Peripheral Shut Down
Most peripherals may be disabled by software in order to cease internal clock generation and remain in a static state. Each peripheral has its own specific disabling sequence (refer to each peripheral description for further details). A peripheral may be disabled at any time and will remain disabled during any low-power mode of operation.
9.4.2
Peripheral Behavior in Low-Power Modes
9.4.2.1
ColdFire Core
The ColdFire core is disabled during any low-power mode. No recovery time is required when exiting any low-power mode.
9.4.2.2
Static Random-Access Memory (SRAM)
SRAM is disabled during any low-power mode. No recovery time is required when exiting any low-power mode.
9.4.2.3
Flash
9.4.2.4
The Flash module is in a low-power state if not being accessed. No recovery time is required after exit from any low-power mode. System Control Module (SCM)
The SCM’s core watchdog timer can bring the device out of all low-power modes except stop mode. In stop mode, all clocks stop, and the core watchdog does not operate.
When enabled, the core watchdog can bring the device out of low-power mode via a core watchdog interrupt. This system setup must meet the conditions specified in
Section 9.4.1, “Low-Power Modes
” for the core watchdog interrupt to bring the part out of low-power mode.
9.4.2.5
DMA Controller (DMA0–DMA3)
In wait and doze modes, the DMA controller is capable of bringing the device out of a low-power mode by generating an interrupt either upon completion of a transfer or upon an error condition. The completion of transfer interrupt is generated when DMA interrupts are enabled by the setting of the DCR[INT] bit, and an interrupt is generated when the DSR[DONE] bit is set. The interrupt upon error condition is
9-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Power Management generated when the DCR[INT] bit is set, and an interrupt is generated when either the CE, BES or BED bit in the DSR becomes set.
The DMA controller is stopped in stop mode and thus cannot cause an exit from this low-power mode.
9.4.2.6
UART Modules (UART0, UART1, and UART2)
In wait and doze modes, the UART may generate an interrupt to exit the low-power modes.
• Clearing the transmit enable bit (TE) or the receiver enable bit (RE) disables UART functions.
• The UARTs are unaffected by wait mode and may generate an interrupt to exit this mode.
In stop mode, the UARTs stop immediately and freeze their operation, register values, state machines, and external pins. During this mode, the UART clocks are shut down. Coming out of stop mode returns the
UARTs to operation from the state prior to the low-power mode entry.
9.4.2.7
I
2
C Module
When the I mode, the I
2
2
C Module is enabled by the setting of the I2CR[IEN] bit and when the device is not in stop
C module is operable and may generate an interrupt to bring the device out of a low-power mode. For an interrupt to occur, the I2CR[IIE] bit must be set to enable interrupts, and the setting of the
I2SR[IIF] generates the interrupt signal to the CPU and interrupt controller. The setting of I2SR[IIF] signifies either the completion of one byte transfer or the reception of a calling address matching its own specified address when in slave receive mode.
In stop mode, the I
2
C Module stops immediately and freezes operation, register values, and external pins.
Upon exiting stop mode, the I
2
C resumes operation unless stop mode was exited by reset.
9.4.2.8
Queued Serial Peripheral Interface (QSPI)
In wait and doze modes, the queued serial peripheral interface (QSPI) may generate an interrupt to exit the low-power modes.
• Clearing the QSPI enable bit (SPE) disables the QSPI function.
• The QSPI is unaffected by wait mode and may generate an interrupt to exit this mode.
In stop mode, the QSPI stops immediately and freezes operation, register values, state machines, and external pins. During this mode, the QSPI clocks are shut down. Coming out of stop mode returns the QSPI to operation from the state prior to the low-power mode entry.
9.4.2.9
DMA Timers (DTIM0–DTIM3)
In wait and doze modes, the DMA timers may generate an interrupt to exit a low-power mode. This interrupt can be generated when the DMA Timer is in either input capture mode or reference compare mode.
In input capture mode, where the capture enable (CE) field of the timer mode register (DTMR) has a non-zero value and the DMA enable (DMAEN) bit of the DMA timer extended mode register (DTXMR) is cleared, an interrupt is issued upon a captured input. In reference compare mode, where the output
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 9-13
Power Management reference request interrupt enable (ORRI) bit of DTMR is set and the DTXMR[DMAEN] bit is cleared, an interrupt is issued when the timer counter reaches the reference value.
DMA timer operation is disabled in stop mode, but the DMA timer is unaffected by either the wait or doze modes and may generate an interrupt to exit these modes. Upon exiting stop mode, the timer will resume operation unless stop mode was exited by reset.
9.4.2.10
Interrupt Controllers (INTC0, INTC1)
The interrupt controller is not affected by any of the low-power modes. All logic between the input sources and generating the interrupt to the processor will be combinational to allow the ability to wake up the CPU processor during low-power stop mode when all system clocks are stopped.
An interrupt request will cause the CPU to exit a low-power mode only if that interrupt’s priority level is at or above the level programmed in the interrupt priority mask field of the CPU’s status register (SR). The interrupt must also be enabled in the interrupt controller’s interrupt mask register as well as at the module from which the interrupt request would originate.
9.4.2.11
I/O Ports
The I/O ports are unaffected by entry into a low-power mode. These pins may impact low-power current draw if they are configured as outputs and are sourcing current to an external load. If low-power mode is exited by a reset, the state of the I/O pins will revert to their default direction settings.
9.4.2.12
Reset Controller
A power-on reset (POR) will always cause a chip reset and exit from any low-power mode.
In wait and doze modes, asserting the external RESET pin for at least four clocks will cause an external reset that will reset the chip and exit any low-power modes.
In stop mode, the RESET pin synchronization is disabled and asserting the external RESET pin will asynchronously generate an internal reset and exit any low-power modes. Registers will lose current values and must be reconfigured from reset state if needed.
If the phase lock loop (PLL) in the clock module is active and if the appropriate (LOCRE, LOLRE) bits in the synthesizer control register are set, then any loss-of-clock or loss-of-lock will reset the chip and exit any low-power modes.
If the watchdog timer is still enabled during wait or doze modes, then a watchdog timer timeout may generate a reset to exit these low-power modes.
When the CPU is inactive, a software reset cannot be generated to exit any low-power mode.
9.4.2.13
Chip Configuration Module
The Chip Configuration Module is unaffected by entry into a low-power mode. If low-power mode is exited by a reset, chip configuration may be executed if configured to do so.
9-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Power Management
9.4.2.14
Clock Module
In wait and doze modes, the clocks to the CPU, Flash, and SRAM will be stopped and the system clocks to the peripherals are enabled. Each module may disable the module clocks locally at the module level. In stop mode, all clocks to the system will be stopped.
During stop mode, the PLL continues to run. The external CLKOUT signal may be enabled or disabled when the device enters stop mode, depending on the LPCR[STPMD] bit settings.The external CLKOUT output pin may be disabled to lower power consumption via the SYNCR[DISCLK] bit. The external
CLKOUT pin function is enabled by default at reset.
9.4.2.15
Edge Port
In wait and doze modes, the edge port continues to operate normally and may be configured to generate interrupts (either an edge transition or low level on an external pin) to exit the low-power modes.
In stop mode, there is no system clock available to perform the edge detect function. Thus, only the level detect logic is active (if configured) to allow any low level on the external interrupt pin to generate an interrupt (if enabled) to exit the stop mode.
9.4.2.16
Programmable Interrupt Timers (PIT0–PIT1)
In stop mode (or in doze mode, if so programmed), the programmable interrupt timer (PIT) ceases operation, and freezes at the current value. When exiting these modes, the PIT resumes operation from the stopped value. It is the responsibility of software to avoid erroneous operation.
When not stopped, the PIT may generate an interrupt to exit the low-power modes.
9.4.2.17
FlexCAN
When enabled, the FlexCAN module is capable of generating interrupts and bringing the device out of a low-power mode. The module has 35 interrupt sources (32 sources due to message buffers and 3 sources due to Bus-off, Error and Wake-up).
When in stop mode, a recessive to dominant transition on the CAN bus causes the WAKE-INT bit in the error & status register to be set. This event can cause a CPU interrupt if the WAKE-MASK bit in module configuration register (MCR) is set.
When setting stop mode in the FlexCAN (by setting the MCR[STOP] bit), the FlexCAN checks for the
CAN bus to be either idle or waits for the third bit of intermission and checks to see if it is recessive. When this condition exists, the FlexCAN waits for all internal activity other than in the CAN bus interface to complete and then the following occurs:
• The FlexCAN shuts down its clocks, stopping most of the internal circuits, to achieve maximum possible power saving.
• The internal bus interface logic continues operation, enabling CPU to access the MCR register.
• The FlexCAN ignores its Rx input pin, and drives its Tx pins as recessive.
• FlexCAN loses synchronization with the CAN bus, and STOP_ACK and NOT_RDY bits in MCR register are set.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 9-15
Power Management
Exiting stop mode is done in one of the following ways:
• Reset the FlexCAN (either by hard reset or by asserting the SOFT_RST bit in MCR).
• Clearing the STOP bit in the MCR.
• Self-wake mechanism. If the SELF-WAKE bit in the MCR is set at the time the FlexCAN enters stop mode, then upon detection of recessive to dominant transition on the CAN bus, the FlexCAN resets the STOP bit in the MCR and resumes its clocks.
Recommendations for, and features of, FlexCAN’s stop mode operation are as follows:
• Upon stop/self-wake mode entry, the FlexCAN tries to receive the frame that caused it to wake; that is, it assumes that the dominant bit detected is a start-of-frame bit. It does not arbitrate for the
CAN bus then.
• Before asserting stop Mode, the CPU should disable all interrupts in the FlexCAN, otherwise it may be interrupted while in stop mode upon a non-wake-up condition. If desired, the
WAKE-MASK bit should be set to enable the WAKE-INT.
• If stop mode is asserted while the FlexCAN is BUSOFF (see error and status register), then the
FlexCAN enters stop mode and stops counting the synchronization sequence; it continues this count once stop mode is exited.
• The correct flow to enter stop mode with SELF-WAKE:
— assert SELF-WAKE at the same time as STOP.
— wait for STOP_ACK bit to be set.
• The correct flow to negate STOP with SELF-WAKE:
— negate SELF-WAKE at the same time as STOP.
— wait for STOP_ACK negation.
• SELF-WAKE should be set only when the MCR[STOP] bit is negated and the FlexCAN is ready; that is, the NOT_RDY bit in the MCR is negated.
• If STOP and SELF_WAKE are set and if a recessive to dominant edge immediately follows on the
CAN bus, the STOP_ACK bit in the MCR may never be set, and the STOP bit in the MCR is reset.
• If the user does not want to have old frames sent when the FlexCAN is awakened (STOP with
Self-Wake), the user should disable all Tx sources, including remote-response, before stop mode entry.
• If halt mode is active at the time the STOP bit is set, then the FlexCAN assumes that halt mode should be exited; hence it tries to synchronize to the CAN bus (11 consecutive recessive bits), and only then does it search for the correct conditions to stop.
• Trying to stop the FlexCAN immediately after reset is allowed only after basic initialization has been performed.
If stop with self-wake is activated, and the FlexCAN operates with single system clock per time-quanta, then there are extreme cases in which FlexCAN's wake-up upon recessive to dominant edge may not conform to the standard CAN protocol, in the sense that the FlexCAN synchronization is shifted one time quanta from the required timing. This shift lasts until the next recessive to dominant edge, which re-synchronizes the FlexCAN back to conform to the protocol. The same holds for auto-power save mode upon wake-up by recessive to dominant edge.
9-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Power Management
The auto-power save mode in the FlexCAN is intended to enable NORMAL operation with optimized power saving. Upon setting the AUTO POWER SAVE bit in the MCR register, the FlexCAN looks for a set of conditions in which there is no need for clocks to run. If all these conditions are met, then the
FlexCAN stops its clocks, thus saving power. While its clocks are stopped, if any of the conditions below is not met, the FlexCAN resumes its clocks. It then continues to monitor the conditions and stops/resumes its clocks appropriately.
The following are conditions for the automatic shut-off of FlexCAN clocks:
• No Rx/Tx frame in progress.
• No moving of Rx/Tx frames between SMB and MB and no Tx frame is pending for transmission in any MB.
• No host access to the FlexCAN module.
• The FlexCAN is neither in halt mode (MCR bit 8), in stop mode (MCT bit 15), nor in BUSOFF.
9.4.2.18
PWM Module
The PWM module is user programmable as to how it behaves when the device enters wait mode
(PWMCTL[PSWAI]) and doze mode (PWMCTL[PFRZ]). If either of these bits are set the PWM input clock to the prescalar will be disabled during the respective low power mode.
In stop mode the input clock is disabled and PWM generation is halted.
9.4.2.19
BDM
Entering halt mode via the BDM port (by asserting the external BKPT pin) will cause the CPU to exit any low-power mode.
9.4.2.20
JTAG
The JTAG (Joint Test Action Group) controller logic is clocked using the TCLK input and is not affected by the system clock. The JTAG cannot generate an event to cause the CPU to exit any low-power mode.
Toggling TCLK during any low-power mode will increase the system current consumption.
9.4.3
Summary of Peripheral State During Low-Power Modes
The functionality of each of the peripherals and CPU during the various low-power modes is summarized
automatically assumes when the STOP instruction is executed and the LPCR[LPMD] field is set for the particular low-power mode. Individual peripherals may be disabled by programming its dedicated control bits. The wakeup capability field refers to the ability of an interrupt or reset by that peripheral to force the
CPU into run mode.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 9-17
Power Management
Module
Table 9-11. CPU and Peripherals in Low-Power Modes
Peripheral Status
1
/ Wakeup Capability
Wait Mode Doze Mode Stop Mode
CPU
SRAM
Flash
System Control Module
Random Number Generator Accelerator
DMA Controller
UART0, UART1 and UART2
I
2
C Module
Stopped
Stopped
Stopped
Enabled
Enabled
Enabled
Enabled
Enabled
No
No
No
Yes
3
No
Yes
Yes
2
Yes
2
Yes
2
Yes
2
Yes
2
Stopped
Stopped
Stopped
Enabled
Stopped
Enabled
Enabled
Enabled
No
No
No
Yes
3
No
Yes
Yes
2
Yes
2
Yes
2
Yes
2
Yes
2
Stopped
Stopped
Stopped
Stopped
Stopped
Stopped
Stopped
Stopped
No
No
No
No
No
No
No
No
QSPI
DMA Timers
Interrupt Controller
I/O Ports
Reset Controller
Chip Configuration Module
Enabled
Enabled
Enabled
Enabled
Enabled
Enabled
No
Yes
3
No
Enabled
Enabled
Enabled
Enabled
Enabled
Enabled
No
Yes
3
No
Stopped
Stopped
Enabled
Enabled
Enabled
Stopped
No
No
Yes
2
No
Yes
No
3
Power Management
Clock Module
Real-Time Clock Module
Edge port
Programmable Interrupt Timers
ADC
General Purpose Timer
FlexCAN
Enabled
Enabled
Enabled
Enabled
Enabled
Enabled
Enabled
Enabled
Yes
2
Yes
2
Yes
2
Yes
2
No
Yes
2
Yes
2
Yes
2
No
Yes
4
Enabled
Enabled
Enabled
Enabled
Program
Program
Enabled
Enabled
Yes
2
Yes
2
Yes
2
Yes
2
No
Yes
2
Yes
2
Yes
2
No
Yes
4
Stopped
Enabled
Enabled
Stopped
Stopped
Stopped
Stopped
Stopped
No
Yes
???
Yes
No
No
No
No
2
2
PWM
BDM
Program
Enabled
Program
Enabled
Stopped
Enabled
No
Yes
4
JTAG Enabled No Enabled No Enabled No
3
4
1
2
“Program” Indicates that the peripheral function during the low-power mode is dependent on programmable bits in the peripheral register map.
These modules can generate a interrupt which will exit a low-power mode. The CPU will begin to service the interrupt exception after wakeup.
These modules can generate a reset which will exit any low-power mode.
The BDM logic is clocked by a separate TCLK clock. Entering halt mode via the BDM port exits any low-power mode.
Upon exit from halt mode, the previous low-power mode will be re-entered and changes made in halt mode will remain in effect.
9-18
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Power Management
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 9-19
Power Management
9-20
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Chapter 10
Reset Controller Module
10.1
Introduction
The reset controller is provided to determine the cause of reset, assert the appropriate reset signals to the system, and keep a history of what caused the reset. The low voltage detection module, which generates low-voltage detect (LVD) interrupts and resets, is implemented within the reset controller module.
10.2
Features
Module features include the following:
• Seven sources of reset:
— External reset input
— Power-on reset (POR)
— Phase locked-loop (PLL) loss of lock
— PLL loss of clock
— Software
— Low-voltage detector (LVD)
— JTAG CLAMP, HIGHZ and EXTEST instructions
• Software-assertable RSTO pin independent of chip reset state
• Software-readable status flags indicating the cause of the last reset
• LVD control and status bits for setup and use of LVD reset or interrupt
10.3
Block Diagram
illustrates the reset controller and is explained in the following sections.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
10-1
Reset Controller Module
RSTI
Pin
Power-On
Reset
PLL
Loss of Clock
PLL
Loss of Lock
Software
Reset
LVD
Detect
Reset
Controller
RSTO
Pin
To Internal Resets
Figure 10-1. Reset Controller Block Diagram
10.4
Signals
provides a summary of the reset controller signal properties. The signals are described in the following sections.
Table 10-1. Reset Controller Signal Properties
Name Direction
Input
Hysteresis
RSTI I Yes
RSTO O —
1
RSTI is always synchronized except when in low-power stop mode.
Input
Synchronization
Yes
1
—
10.4.1
RSTI
Asserting the external RSTI for at least four rising CLKOUT edges causes the external reset request to be recognized and latched.
10.4.2
RSTO
This active-low output signal is driven low when the internal reset controller module resets the chip. When
RSTO is active, the user can drive override options on the data bus.
10.5
Memory Map and Registers
The reset controller programming model consists of these registers:
• Reset control register (RCR)—selects reset controller functions
• Reset status register (RSR)—reflects the state of the last reset source
10-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Reset Controller Module
See Table 10-2 for the memory map and the following paragraphs for a description of the registers.
Table 10-2. Reset Controller Memory Map
IPSBAR
Offset
1
Register
0x11_0000 Reset Control Register (RCR)
Width
(bits)
Access Reset Value Section/Page
8 R/W
0x11_0001 Reset Status Register (RSR) 8 R
1
Addresses not assigned to a register and undefined register bits are reserved for expansion.
0x05
10.5.1
Reset Control Register (RCR)
The RCR allows software control for requesting a reset, independently asserting the external RSTO pin, and controlling low-voltage detect (LVD) functions.
Access: User read/write IPSBAR
Offset:
0x11_0000 (RCR)
7
R
W
Reset:
SOFTRST
0
6
FRCRSTO
UT
5
0
4
LVDF
3
LVDIE
2
LVDRE
0 0 0 0
Figure 10-2. Reset Control Register (RCR)
1
1
0
0
0
LVDE
1
Table 10-3. RCR Field Descriptions
Field Description
7
SOFTRST
Allows software to request a reset. The reset caused by setting this bit clears this bit.
1 Software reset request
0 No software reset request
6
FRCRSTOUT
Allows software to assert or negate the external RSTO pin.
1 Assert RSTO pin
0 Negate RSTO pin
CAUTION: External logic driving reset configuration data during reset needs to be considered when asserting the RSTO pin when setting FRCRSTOUT.
Reserved, should be cleared.
5
—
4
LVDF
LVD flag. Indicates the low-voltage detect status if LVDE is set. Write a 1 to clear the LVDF bit.
1 Low voltage has been detected
0 Low voltage has not been detected
NOTE: The setting of this flag causes an LVD interrupt if LVDE and LVDIE bits are set and LVDRE is cleared when the supply voltage V
DD
drops below V
DD
(minimum). The vector for this interrupt is shared with INT0 of the EPORT module. Interrupt arbitration in the interrupt service routine is necessary if both of these interrupts are enabled. Also, LVDF is not cleared at reset, however it will always initialize to a zero since the part will not come out of reset while in a low-power state (LVDE/LVDRE bits are enabled out of reset).
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
10-3
Reset Controller Module
Field
3
LVDIE
2
LVDRE
1
—
0
LVDE
Table 10-3. RCR Field Descriptions (continued)
Description
LVD interrupt enable. Controls the LVD interrupt if LVDE is set. This bit has no effect if the LVDE bit is a logic 0.
1 LVD interrupt enabled
0 LVD interrupt disabled
LVD reset enable. Controls the LVD reset if LVDE is set. This bit has no effect if the LVDE bit is a logic 0. LVD reset has priority over LVD interrupt, if both are enabled.
1 LVD reset enabled
0 LVD reset disabled
Reserved, should be cleared.
Controls whether the LVD is enabled.
1 LVD is enabled
0 LVD is disabled
10.5.2
Reset Status Register (RSR)
The RSR contains a status bit for every reset source. When reset is entered, the cause of the reset condition is latched, along with a value of 0 for the other reset sources that were not pending at the time of the reset condition. These values are then reflected in RSR. One or more status bits may be set at the same time.
The cause of any subsequent reset is also recorded in the register, overwriting status from the previous reset condition.
RSR can be read at any time. Writing to RSR has no effect.
Access: User read-only IPSBAR
Offset:
0x11_0001 (RSR)
R
W
7
0
Reset: Reset Dependent
6
LVD
5
SOFT
4
0
3
POR
Figure 10-3. Reset Status Register (RSR)
2
EXT
1
LOC
0
LOL
Field
7
—
6
LVD
Table 10-4. RSR Field Descriptions
Reserved, should be cleared.
Description
5
SOFT
Low voltage detect. Indicates that the last reset state was caused by an LVD reset.
1 Last reset state was caused by an LVD reset
0 Last reset state was not caused by an LVD reset
Software reset flag. Indicates that the last reset was caused by software.
1 Last reset caused by software
0 Last reset not caused by software
10-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Field
4
—
3
POR
2
EXT
1
LOC
0
LOL
Reset Controller Module
Table 10-4. RSR Field Descriptions (continued)
Description
Reserved, should be cleared.
Power-on reset flag. Indicates that the last reset was caused by a power-on reset.
1 Last reset caused by power-on reset
0 Last reset not caused by power-on reset
External reset flag. Indicates that the last reset was caused by an external device asserting the external RSTI pin.
1 Last reset state caused by external reset
0 Last reset not caused by external reset
Loss-of-clock reset flag. Indicates that the last reset state was caused by a PLL loss of clock.
1 Last reset caused by loss of clock
0 Last reset not caused by loss of clock
Loss-of-lock reset flag. Indicates that the last reset state was caused by a PLL loss of lock.
1 Last reset caused by a loss of lock
0 Last reset not caused by loss of lock
10.6
Functional Description
10.6.1
Reset Sources
defines the sources of reset and the signals driven by the reset controller.
Table 10-5. Reset Source Summary
Type Source
Power on
External RSTI pin (not stop mode)
External RSTI pin (during stop mode)
Loss-of-clock
Loss-of-lock
Software
LVD reset
Asynchronous
Synchronous
Asynchronous
Asynchronous
Asynchronous
Synchronous
Asynchronous
To protect data integrity, a synchronous reset source is not acted upon by the reset control logic until the end of the current bus cycle. Reset is then asserted on the next rising edge of the system clock after the cycle is terminated. Whenever the reset control logic must synchronize reset to the end of the bus cycle, the internal bus monitor is automatically enabled regardless of the BME bit state in the chip configuration register (CCR). Then, if the current bus cycle is not terminated normally, the bus monitor terminates the cycle based on the length of time programmed in the BMT field of the CCR.
Internal byte, word, or longword writes are guaranteed to complete without data corruption when a synchronous reset occurs. External writes, including longword writes to 16-bit ports, are also guaranteed to complete.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
10-5
Reset Controller Module
Asynchronous reset sources usually indicate a catastrophic failure. Therefore, the reset control logic does not wait for the current bus cycle to complete. Reset is asserted immediately to the system.
10.6.1.1
Power-On Reset
At power up, the reset controller asserts RSTO. RSTO continues to be asserted until V
DD
has reached a minimum acceptable level and, if PLL clock mode is selected, until the PLL achieves phase lock. Then after approximately another 512 cycles, RSTO is negated and the part begins operation.
10.6.1.2
External Reset
Asserting the external RSTI for at least four rising CLKOUT edges causes the external reset request to be recognized and latched. The bus monitor is enabled and the current bus cycle is completed. The reset controller asserts RSTO for approximately 512 cycles after RSTI is negated and the PLL has acquired lock.
The part then exits reset and begins operation.
In low-power stop mode, the system clocks are stopped. Asserting the external RSTI in stop mode causes an external reset to be recognized.
10.6.1.3
Loss-of-Clock Reset
This reset condition occurs in PLL clock mode when the LOCRE bit in the SYNCR is set and either the
PLL reference or the PLL itself fails. The reset controller asserts RSTO for approximately 512 cycles after the PLL has acquired lock. The device then exits reset and begins operation.
10.6.1.4
Loss-of-Lock Reset
This reset condition occurs in PLL clock mode when the LOLRE bit in the SYNCR is set and the PLL loses lock. The reset controller asserts RSTO for approximately 512 cycles after the PLL has acquired lock. The device then exits reset and resumes operation.
10.6.1.5
Software Reset
A software reset occurs when the SOFTRST bit is set. If the RSTI is negated and the PLL has acquired lock, the reset controller asserts RSTO for approximately 512 cycles. Then the device exits reset and resumes operation.
10.6.1.6
LVD Reset
The LVD reset will occur when the supply input voltage, V
DD, drops below V
LVD
(minimum).
10.6.2
Reset Control Flow
The reset logic control flow is shown in
. In this figure, the control state boxes have been numbered, and these numbers are referred to (within parentheses) in the flow description that follows. All cycle counts given are approximate.
10-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Reset Controller Module
0 POR OR LVD
12
NEGATE RSTO
1
LOSS OF CLOCK?
Y
N
2
LOSS OF LOCK?
Y
N
3 RSTI
PIN OR WD TIMEOUT
OR SW RESET?
Y
N
5
ENABLE BUS MONITOR
6
BUS CYCLE
COMPLETE?
Y
7
ASSERT RSTO AND
LATCH RESET STATUS
N
8
RSTI NEGATED?
N
Y
9
PLL MODE?
N
10
WAIT 512 CLKOUT CYCLES
Y
11
RCON ASSERTED?
Y
N
Figure 10-4. Reset Control Flow
4
ASSERT RSTO AND
LATCH RESET STATUS
9A
PLL LOCKED?
Y
N
11A
LATCH CONFIGURATION
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
10-7
Reset Controller Module
10.6.2.1
Synchronous Reset Requests
In this discussion, the references in parentheses refer to the state numbers in
. All cycle counts given are approximate.
If the external RSTI signal is asserted by an external device for at least four rising CLKOUT edges (3) and if software requests a reset, the reset control logic latches the reset request internally and enables the bus monitor (5). When the current bus cycle is completed (6), RSTO is asserted (7). The reset control logic waits until the RSTI signal is negated (8) and for the PLL to attain lock (9, 9A) before waiting 512
CLKOUT cycles (1). The reset control logic may latch the configuration according to the RCON signal level (11, 11A) before negating RSTO (12).
If the external RSTI signal is asserted by an external device for at least four rising CLKOUT edges during the 512 count (10) or during the wait for PLL lock (9A), the reset flow switches to (8) and waits for the
RSTI signal to be negated before continuing.
10.6.2.2
Internal Reset Request
If reset is asserted by an asynchronous internal reset source, such as loss of clock (1) or loss of lock (2), the reset control logic asserts RSTO (4). The reset control logic waits for the PLL to attain lock (9, 9A) before waiting 512 CLKOUT cycles (1). Then the reset control logic may latch the configuration according to the RCON pin level (11, 11A) before negating RSTO (12).
If loss of lock occurs during the 512 count (10), the reset flow switches to (9A) and waits for the PLL to lock before continuing.
10.6.2.3
Power-On Reset/Low-Voltage Detect Reset
When the reset sequence is initiated by power-on reset (0), the same reset sequence is followed as for the other asynchronous reset sources.
10.6.3
Concurrent Resets
This section describes the concurrent resets. As in the previous discussion, references in parentheses refer
to the state numbers in Figure 10-4 .
10.6.3.1
Reset Flow
If a power-on reset or low-voltage detect condition is detected during any reset sequence, the reset sequence starts immediately (0).
If the external RSTI pin is asserted for at least four rising CLKOUT edges while waiting for PLL lock or the 512 cycles, the external reset is recognized. Reset processing switches to wait for the external RSTI pin to negate (8).
If a loss-of-clock or loss-of-lock condition is detected while waiting for the current bus cycle to complete
(5, 6) for an external reset request, the cycle is terminated. The reset status bits are latched (7) and reset processing waits for the external RSTI pin to negate (8).
10-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Reset Controller Module
If a loss-of-clock or loss-of-lock condition is detected during the 512 cycle wait, the reset sequence continues after a PLL lock (9, 9A).
10.6.3.2
Reset Status Flags
For a POR reset, the POR and LVD bits in the RSR are set, and the SOFT, WDR, EXT, LOC, and LOL bits are cleared even if another type of reset condition is detected during the reset sequence for the POR.
If a loss-of-clock or loss-of-lock condition is detected while waiting for the current bus cycle to complete
(5, 6) for an external reset request, the EXT, SOFT, and/or WDR bits along with the LOC and/or LOL bits are set.
If the RSR bits are latched (7) during the EXT, SOFT, and/or WDR reset sequence with no other reset conditions detected, only the EXT, SOFT, and/or WDR bits are set.
If the RSR bits are latched (4) during the internal reset sequence with the RSTI pin not asserted and no
SOFT or WDR event, then the LOC and/or LOL bits are the only bits set.
For a LVD reset, the LVD bit in the RSR is set, and the SOFT, WDR, EXT, LOC, and LOL bits are cleared to 0, even if another type of reset condition is detected during the reset sequence for LVD.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
10-9
Reset Controller Module
10-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 11
Static RAM (SRAM)
11.1
Introduction
This chapter describes the on-chip static RAM (SRAM) implementation, including general operations, configuration, and initialization. It also provides information and examples showing how to minimize power consumption when using the SRAM.
11.1.1
Features
The major features include the following:
• One 256-Kbyte SRAM (128 Kbyte for MCF52230 and MCF52231)
• Single-cycle access
• Physically located on the processor's high-speed local bus
• Memory location programmable on any 0-modulo- Kbyte address
• Byte, word, and longword address capabilities
11.1.2
Operation
The SRAM module provides a general-purpose memory block that the ColdFire processor can access in a single cycle. The location of the memory block can be specified to any 0-modulo-K address within the
256-byte address space. The memory is ideal for storing critical code or data structures or for use as the system stack. Because the SRAM module is physically connected to the processor's high-speed local bus, it can service processor-initiated accesses or memory-referencing commands from the debug module.
The SRAM is dual-ported to provide access. The SRAM is partitioned into two physical memory arrays to allow simultaneous access to both arrays by the processor core and another bus master. See Chapter 10,
“System Control Module (SCM),” for more information.
11.2
Memory Map/Register Description
The SRAM programming model, shown in
, includes a description of the SRAM base address register (RAMBAR), SRAM initialization, and power management.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
11-1
Static RAM (SRAM)
Table 11-1. SRAM Programming Model
CPU Space
(Rc)
0xC05
Register
Width
(bits)
Access Reset Value
Written w/ MOVEC
Section/Page
Supervisor Access Only Registers
RAM Base Address Register (RAMBAR) 32 R/W 0x0000_0000 Yes
11.2.1
SRAM Base Address Register (RAMBAR)
The configuration information in the SRAM base address register (RAMBAR) controls the operation of the SRAM module.
• The RAMBAR holds the base address of the SRAM. The MOVEC instruction provides write-only access to this register.
• The RAMBAR can be read or written from the debug module.
• All undefined bits in the register are reserved. These bits are ignored during writes to the
RAMBAR and return zeroes when read from the debug module.
• The RAMBAR valid bit is cleared by reset, disabling the SRAM module. All other bits are unaffected.
Section 10.5.2, “Memory Base Address Register (RAMBAR).” The RAMBAR contains several control
fields. These fields are shown in Figure 11-1
.
Address: CPU @ 0x0C05 Access: Core write-only
Debug read/write
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11
R 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Reset 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
10
0
0
9 8 7 6 5 4 3 2 1 0
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
Figure 11-1. SRAM Base Address Register (RAMBAR)
Field
31–
BA
–12
Table 11-2. RAMBAR Field Descriptions
Description
Base address. Defines the 0-modulo-K base address of the SRAM module (0-modulo-16 for the MCF5211).
By programming this field, the SRAM may be located on any -Kbyte boundary within the processor’s byte address space (16-Kbyte boundary for the MCF5211). Bit 14 is reserved on the MCF5212 and MCF5213 devices and should be cleared.
Reserved, should be cleared.
11-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Static RAM (SRAM)
Table 11-2. RAMBAR Field Descriptions (continued)
Field
11–10
PRIU
PRIL
Description
Priority bit. PRIU determines if or CPU has priority in the upper K bank of memory. PRIL determines if or
CPU has priority in the lower K bank of memory. If a bit is set, the CPU has priority. If a bit is cleared, has priority. Priority is determined according to the following table:
PRIU,PRIL Upper Bank Priority Lower Bank Priority
00
01
10
11
CPU
CPU
CPU
CPU
Note: The recommended setting (maximum performance) for the priority bits is 00.
8
WP
Write protect. Allows only read accesses to the SRAM. When this bit is set, any attempted write access from the core will generate an access error exception to the ColdFire processor core.
0 Allows core read and write accesses to the SRAM module
1 Allows only core read accesses to the SRAM module
Note: Non-core write accesses are not affected by this bit.
Reserved, should be cleared.
7
5–1
C/I, SC, SD, UC,
UD
Address space masks (AS n ). These five bit fields allow certain types of accesses to be “masked,” or inhibited from accessing the SRAM module. The address space mask bits are:
C/I = CPU space/interrupt acknowledge cycle mask
SC = Supervisor code address space mask
SD = Supervisor data address space mask
UC = User code address space mask
UD = User data address space mask
0
V
For each address space bit:
0 An access to the SRAM module can occur for this address space
1 Disable this address space from the SRAM module. If a reference using this address space is made, it is inhibited from accessing the SRAM module, and is processed like any other non-SRAM reference.
These bits are useful for power management as detailed in Section 11.3.2, “Power Management.”
In most applications the C/I bit is set
Valid. When set, this bit enables the SRAM module; otherwise, the module is disabled. A hardware reset clears this bit.
11.3
Initialization/Application Information
After a hardware reset, the contents of the SRAM module are undefined. The valid bit of the RAMBAR is cleared, disabling the module. If the SRAM requires initialization with instructions or data, the following steps should be performed:
1. Load the RAMBAR, mapping the SRAM module to the desired location within the address space.
2. Read the source data and write it to the SRAM. There are various instructions to support this function, including memory-to-memory move instructions, or the MOVEM opcode. The MOVEM instruction is optimized to generate line-sized burst fetches on 0-modulo-16 addresses, so this opcode generally provides maximum performance.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
11-3
Static RAM (SRAM)
3. After the data has been loaded into the SRAM, it may be appropriate to load a revised value into the RAMBAR with a new set of attributes. These attributes consist of the write-protect and address space mask fields.
The ColdFire processor or an external debugger using the debug module can perform these initialization functions.
11.3.1
SRAM Initialization Code
The following code segment describes how to initialize the SRAM. The code sets the base address of the
SRAM at 0x000_0000 and initializes the SRAM to zeros.
RAMBASE EQU 0x0000000 ;set this variable to 0x0000000 move.l movec.l
#RAMBASE+RAMVALID,D0
D0, RAMBAR
;load RAMBASE + valid bit into D0.
;load RAMBAR and enable SRAM
The following loop initializes the entire SRAM to zero: lea.l move.l
RAMBASE,A0
#,D0
;load pointer to SRAM
;load loop counter into D0 (SRAM size/4)
SRAM_INIT_LOOP: clr.l subq.l bne.b
(A0)+
#1,D0
SRAM_INIT_LOOP
;clear 4 bytes of SRAM
;decrement loop counter
;if done, then exit; else continue looping
11.3.2
Power Management
If the SRAM is used only for data operands, setting the AS n bits associated with instruction fetches can decrease power dissipation. Additionally, if the SRAM contains only instructions, masking operand accesses can reduce power dissipation.
shows some examples of typical RAMBAR settings.
Table 11-3. Typical RAMBAR Setting Examples
Data Contained in SRAM
Instruction Only
Data Only
Both Instructions and Data
RAMBAR[7:0]
0x2B
0x35
0x21
11-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 12
Chip Configuration Module (CCM)
12.1
Introduction
This chapter describes the various operating configurations of the device. It provides a description of signals used by the CCM and a programming model.
12.1.1
Block Diagram
The chip configuration module (CCM) controls the chip configuration and mode of operation for the
MCF52235.
Clock Mode
Selection
Output Pad
Strength Selection
Chip Configuration Register
Reset Configuration Register
Chip Identification Register
Chip Test Register
Figure 12-1. Chip Configuration Module Block Diagram
12.1.2
Features
The CCM selects the following:
• External clock or phase-lock loop (PLL) mode with internal or external reference
• Output pad drive strength
• Low-power configuration
• Processor status (PSTAT) and processor debug data (DDATA) functions
• BDM or JTAG mode
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
12-1
Chip Configuration Module (CCM)
12.2
External Signal Descriptions
provides an overview of the CCM signals.
Table 12-1. Signal Properties
RCON
Name Function
Reset configuration select
Reset State
Internal weak pull-up device
12.2.1
RCON
If the external RCON pin is asserted during reset, then various chip functions, including the reset configuration pin functions after reset, are configured according to the levels driven onto the external data pins (see
Section 12.4, “Functional Description
”). The internal configuration signals are driven to reflect the levels on the external configuration pins to allow for module configuration.
12.3
Memory Map/Register Definition
This subsection provides a description of the memory map and registers.
12.3.1
Programming Model
The CCM programming model consists of these registers:
• The chip configuration register (CCR) controls the main chip configuration.
• The reset configuration register (RCON) indicates the default chip configuration.
• The chip identification register (CIR) contains a unique part number.
Some control register bits are implemented as write-once bits. These bits are always readable, but once the bit has been written, additional writes have no effect, except during debug and test operations.
Some write-once bits can be read and written while in debug mode. When debug mode is exited, the chip configuration module resumes operation based on the current register values. If a write to a write-once register bit occurs while in debug mode, the register bit remains writable on exit from debug or test mode.
shows the accessibility of write-once bits.
Table 12-2. Write-Once Bits Read/Write Accessibility
Configuration
All configurations
Debug operation
Read/Write Access
Read-always
Write-always
12-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chip Configuration Module (CCM)
12.3.2
Memory Map
Table 12-3. Chip Configuration Module Memory Map
IPSBAR
Offset
1
Register
Width
(bits)
Access Reset Value Section/Page
Supervisor Mode Access Only
0x11_0004 Chip Configuration Register (CCR)
0x11_0007 Low-Power Control Register (LPCR)
2
0x11_0008 Reset Configuration Register (RCON)
16
8
R
R/W
0x0001
0x00
8.3.3.1/8-3
7.2.4.1/7-12
16 R 0x0000
0x11_000A Chip Identification Register (CIR) 16 R 0x2000
0x11_0010 Unimplemented
3
—
1
2
3
Addresses not assigned to a register and undefined register bits are reserved for expansion.
See Chapter 8, “Power Management,” for a description of the LPCR. It is shown here only to warn against accidental writes to this register.
Accessing an unimplemented address has no effect and causes a cycle termination transfer error.
NOTE
To safeguard against unintentionally activating test logic, write 0x0000 to the above reserved location during initialization (immediately after reset) to lock out test features. Setting any bits in the CCR may lead to unpredictable results.
12.3.3
Register Descriptions
The following section describes the CCM registers.
12.3.3.1
Chip Configuration Register (CCR)
IPSBAR
Offset:
0x11_0004 (CCR)
15
R
W
Reset 0
14
0
0
13
0
0
12
0
11
0
10
0
9
0
8
0
0 0 0 0 0 0 0 0
Figure 12-2. Chip Configuration Register (CCR)
0
Access: Supervisor read-only
7 6 5 4 3
0 SZEN PSTEN 0 BME
0
2
0
1
BMT
0
0
1
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
12-3
Chip Configuration Module (CCM)
Field
1–
—
6
SZEN
5
PSTEN
TSIZ[1:0] enable. This read/write bit enables the TSIZ[1:0] function of the external pins.
0 TSIZ[1:0] function disabled.
1 TSIZ[1:0] function enabled.
PST[3:0]/DDATA[3:0] enable. This read/write bit enables the Processor Status (PST) and Debug Data (DDATA) functions of the external pins.
0 PST/DDATA function disabled.
1 PST/DDATA function enabled.
Reserved, should be cleared.
4
—
3
BME
2–0
BMT
Table 12-4. CCR Field Descriptions
Reserved, should be cleared.
Description
Bus monitor enable. This read/write bit enables the bus monitor to operate during external bus cycles.
0 Bus monitor disabled for external bus cycles.
1 Bus monitor enabled for external bus cycles.
shows the read/write accessibility of this write-once bit.
Bus monitor timing. This field selects the timeout period (in system clocks) for the bus monitor.
000 65536
001 32768
010 16384
011 8192
100 4096
101 2048
110 1024
111 512
shows the read/write accessibility of this write-once bit.
12.3.3.2
Reset Configuration Register (RCON)
At reset, RCON determines the default operation of certain chip functions. All default functions defined by the RCON values can only be overridden during reset configuration (see
Configuration ”) if the external RCON pin is asserted. RCON is a read-only register.
Access: Supervisor read-only IPSBAR
Offset:
0x11_0008 (RCON)
15
R
W
Reset 0
14
0
0
13
0
0
12
0
11
0
10
0
9
0
8
0
7
0
6 5 4
0 RLOAD 0
0 0 0 0 0 0 0 0
Figure 12-3. Reset Configuration Register (RCON)
0
3
0
0
2
0
0
1 0
0 MODE
0 0
12-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chip Configuration Module (CCM)
Field
15–6
—
5
RLOAD
4–1
—
0
MODE
Table 12-5. RCON Field Descriptions
Reserved, should be cleared.
Description
Pad driver load. Reflects the default pad driver strength configuration.
0 artial drive strength (This is the default value.)
1 ull drive strength
Reserved, should be cleared.
Chip configuration mode. Reflects the default chip configuration mode.
0
1 .
The default mode can be overridden during reset configuration.
12.3.3.3
Chip Identification Register (CIR)
IPSBAR
Offset:
0x11_000A (CIR)
14 13 15
R
W
Reset X X X
12 11
PIN
10 9 8 7 6 5 4
X X X X X X X X
Figure 12-4. Chip Identification Register (CIR)
X
3
PRN
2
X X
Access: read-only
1 0
X X
Field
15–6
PIN
Table 12-6. CIR Field Description
Description
Part identification number. Contains a unique identification number for the device.
MCF52230 = 0x48
MCF52233 = 0x4A
MCF52234 = 0x4B
MCF52235 = 0x4C
Part revision number. This number is increased by one for each new full-layer mask set of this part. The revision numbers are assigned in chronological order, beginning with zero.
5–0
PRN
12.4
Functional Description
Three functions are defined within the chip configuration module:
• Reset configuration
• Output pad strength configuration
• Clock mode selections
These functions are described in the following sections.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
12-5
Chip Configuration Module (CCM)
12.4.1
Reset Configuration
During reset, the pins for the reset override functions are immediately configured to known states.
shows the states of the external pins while in reset.
Table 12-7. Reset Configuration Pin States During Reset
Pin
Pin
Function
1
I/O
Output
State
Input
State
RCON RCON function for all modes
2
Input — Internal weak pull-up device
1
2
If the external RCON pin is not asserted during reset, pin functions are determined by the default operation mode defined in the RCON register. If the external RCON pin is asserted, pin functions are determined by the override values driven on the external data bus pins.
During reset, the external RCON pin assumes its RCON pin function, but this pin changes to the function defined by the chip operation mode immediately after reset. See
.
If the RCON pin is not asserted during reset, the chip configuration and the reset configuration pin functions after reset are determined by RCON or fixed defaults, regardless of the states of the external data pins. The internal configuration signals are driven to levels specified by the RCON register’s reset state for default module configuration.
If the RCON pin is asserted during reset, then various chip functions, including the reset configuration pin functions after reset, are configured according to the levels driven onto the external data pins (see
). The internal configuration signals are driven to reflect the levels on the external configuration pins to allow for module configuration.
Table 12-8. Configuration During Reset
1
Pin(s) Affected
Default
Configuration
Override Pins in Reset
2,
Function
All output pins RCON[5] = 1 0
1
Partial strength
Full strength
4
1
2
Modifying the default configurations is possible only if the external RCON pin is asserted.
The external reset override circuitry drives the data bus pins with the override values while RSTOUT is asserted. It must stop driving the data bus pins within one CLKOUT cycle after RSTOUT is negated. To prevent contention with the external reset override circuitry, the reset override pins are forced to inputs during reset and do not become outputs until at least one CLKOUT cycle after RSTOUT is negated. RCON must also be negated within one cycle after
RSTOUT is negated.
12.4.2
Output Pad Strength Configuration
Output pad strength is determined during reset configuration.
12.5
Reset
Reset initializes CCM registers to a known startup state as described in
The CCM controls chip configuration at reset as described in Section 12.4,
12-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 13
System Control Module (SCM)
13.1
Introduction
This section details the functionality of the system control module (SCM) that provides the programming model for the system access control unit (SACU), system bus arbiter, 32-bit core watchdog timer (CWT), and system control registers and logic. Specifically, the system control includes the internal peripheral system (IPS) base address register (IPSBAR), the processor’s dual-port RAM base address register
(RAMBAR), and system control registers that include the core watchdog timer control.
13.2
Overview
The SCM provides the control and status for a variety of functions including base addressing and address space masking for both the IPS peripherals and resources (IPSBAR) and the ColdFire core memory spaces
(RAMBAR). The CPU core supports two memory banks, one for the internal SRAM and the other for the internal Flash.
The SACU provides the mechanism needed to implement secure bus transactions to the system address space.
The programming model for the system bus arbitration resides in the SCM. The SCM sources the necessary control signals to the arbiter for bus master management.
The CWT provides a means of preventing system lockup due to uncontrolled software loops via a special software service sequence. If periodic software servicing action does not occur, the CWT times out with a programmed response (system reset or interrupt) to allow recovery or corrective action to be taken.
13.3
Features
The SCM includes these distinctive features:
• IPS base address register (IPSBAR)
— Base address location for 1-Gbyte peripheral space
— User control bits
• Processor-local memory base address register (RAMBAR)
• System control registers
— Core reset status register (CRSR) indicates type of last reset
— Core watchdog service register (CWSR) services watchdog timer
— Core watchdog control register (CWCR) for watchdog timer control
• System bus master arbitration programming model (MPARK)
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
13-1
System Control Module (SCM)
• System access control unit (SACU) programming model
— Master privilege register (MPR)
— Peripheral access control registers (PACRs)
— Grouped peripheral access control registers (GPACR0, GPACR1)
13.4
Memory Map and Register Definition
The memory map for the SCM registers is shown in
Table 13-1 . All the registers in the SCM are
memory-mapped as offsets within the 1-Gbyte IPS address space and accesses are controlled to these registers by the control definitions programmed into the SACU.
Table 13-1. SCM Register Map
IPSBAR
Offset
1
Register
Width
(bits)
Access Reset Value Section/Page
0x00_0000 IPS Base Address Register (IPSBAR)
0x00_0008 Memory Base Address Register (RAMBAR)
0x00_000C Peripheral Power Management Register High (PPMRH)
2
0x00_0010 Core Reset Status Register (CRSR)
0x00_0011 Core Watchdog Control Register (CWCR)
0x00_0012 Low-Power Interrupt Control Register (LPICR)
0x00_0013 Core Watchdog Service Register (CWSR)
0x00_0014 DMA Request Control Register (DMAREQC)
0x00_0018 Peripheral Power Management Register Low (PPMRL)
2
0x00_001C Default Bus Master Park Register (MPARK)
0x00_0020 Master Privilege Register (MPR)
0x00_0021 Peripheral Power Management Set Register (PPMRS)
2
0x00_0022 Peripheral Power Management Clear Register (PPMRC)
2
0x00_0023 IPS Bus Timeout Monitor Register (IPSBMT)
2,3
0x00_0024 Peripheral Access Control Register (PACR0)
0x00_0025 Peripheral Access Control Register (PACR1)
0x00_0026 Peripheral Access Control Register (PACR2)
0x00_0027 Peripheral Access Control Register (PACR3)
0x00_0028 Peripheral Access Control Register (PACR4)
0x00_002A Peripheral Access Control Register (PACR5)
0x00_002B Peripheral Access Control Register (PACR6)
0x00_002C Peripheral Access Control Register (PACR7)
32
32
8
8
32
32
8
8
8
32
8
8
32
32
32
8
8
8
8
8
8
8
R/W
R/W
R/W
R/W
R/W
R/W
R/W 0x40000001
R/W 0x00000000
R/W 0x00000000
R/W
7.2.1/7-2
R/W
R/W
0x00
0x00
7.2.2/7-9
R/W
R/W 0x00000000 14.4.1/14-5
R/W 0x00000001 7.2.1.1/7-5
R/W 0x30E10000
R/W
W
0x03
0x00
7.2.3/7-11
R/W 0x00
R/W 0x00000008
R/W
R/W
0x00
0x00
7.2.4/7-11
7.3/7-13
0x00
0x00
0x00
0x00
0x00
0x00
13-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
System Control Module (SCM)
Table 13-1. SCM Register Map (continued)
IPSBAR
Offset
1
Register
Width
(bits)
Access Reset Value Section/Page
0x00_002E Peripheral Access Control Register (PACR8)
0x00_0030 GPACR0 Register
8
8
R/W
R/W
0x00
0x00
0x00_0031 GPACR1 Register 8 R/W 0x00
2
3
1
Addresses not assigned to a register and undefined register bits are reserved for expansion.
The PPMRH, LPICR, PMRL, PPMRS, PPMRC, and IPSBMT are described in Chapter 7, “Power Management.”
Register must be addressed as a byte.
Table 13-2.
IPSBAR
Offset
[31:24] [23:16] [15:8] [7:0]
0x00_0000
0x00_0004
IPSBAR
—
0x00_0008
0x00_000C
0x00_0010
0x00_0014
CRSR
RAMBAR
PPMRH
1
CWCR LPICR
1
DMAREQC
2
PPMRL
1
CWSR
0x00_0018
0x00_001C MPARK
2
3
0x00_0020
0x00_0024
0x00_0028
0x00_002c
0x00_0030
0x00_0034
1
0x00_0038
0x00_003C
MPR
PACR0
PACR4
PACR8
GPACR0
—
—
—
PPMRS
PACR1
PACR5
—
GPACR1
—
—
—
1
PPMRC
PACR2
PACR6
—
—
—
—
—
1
IPSBMT
PACR3
PACR7
—
—
—
—
—
The LPICR is described in Chapter 7, “Power Management.”
The DMAREQC register is described in Chapter 14, “DMA Controller Module .”
Register must be addressed as a byte.
1,3
13.5
Register Descriptions
13.5.1
Internal Peripheral System Base Address Register (IPSBAR)
The IPSBAR specifies the base address for the 1-Gbyte memory space associated with the on-chip peripherals. At reset, the base address is loaded with a default location of 0x4000_0000 and marked as valid (IPSBAR[V]=1). If desired, the address space associated with the internal modules can be moved by loading a different value into the IPSBAR at a later time.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
13-3
System Control Module (SCM)
NOTE
Accessing reserved IPSBAR memory space could result in an unterminated bus cycle that causes the core to hang. Only a hard reset will allow the core to recover from this state. Therefore, all bus accesses to IPSBAR space should fall within a module’s memory map space.
If an address hits in overlapping memory regions, the following priority is used to determine what memory is accessed:
1. IPSBAR
2. RAMBAR
NOTE
This is the list of memory access priorities when viewed from the processor core.
See Figure 13-1 and Table 13-3
for descriptions of the bits in IPSBAR.
IPSBAR
Offset:
0x000 (IPSBAR)
31 30
R
BA31 BA30
W
Reset 0 1
29
0
0
28
0
0
27
0
0
26
0
0
25
0
0
24
0
0
23
0
0
22
0
0
21
0
0
20
0
0
19
0
0
Access: read/write
18
0
0
17
0
0
16
0
0
15
R 0
W
Reset 0
14
0
0
13
0
0
12
0
11
0
10
0
9
0
8
0
7
0
6
0
5
0
4
0
3
0
0
2
0
0
1
0
0
0
V
1 0 0 0 0 0 0 0 0
Figure 13-1. IPS Base Address Register (IPSBAR)
0
Table 13-3. IPSBAR Field Description
Field Description
31–30
BA
Base address. Defines the base address of the 1-Gbyte internal peripheral space. This is the starting address for the
IPS registers when the valid bit is set.
29–1 Reserved, should be cleared.
0
V
Valid. Enables/disables the IPS Base address region. V is set at reset.
0 IPS Base address is not valid.
1 IPS Base address is valid.
13.5.2
Memory Base Address Register (RAMBAR)
The device supports dual-ported local SRAM memory. This processor-local memory can be accessed directly by the core and/or other system bus masters. Since this memory provides single-cycle accesses at processor speed, it is ideal for applications where double-buffer schemes can be used to maximize
13-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
System Control Module (SCM) system-level performance. For example, a DMA channel in a typical double-buffer application (also known as a ping-pong scheme) may load data into one portion of the dual-ported SRAM while the processor is manipulating data in another portion of the SRAM. Once the processor completes the data calculations, it begins processing the just-loaded buffer while the DMA moves out the just-calculated data from the other buffer, and reloads the next data block into the just-freed memory region. The process repeats with the processor and the DMA ping-ponging between alternate regions of the dual-ported
SRAM.
The device design implements the dual-ported SRAM in the memory space defined by the RAMBAR register. There are two physical copies of the RAMBAR register: one located in the processor core and accessible only via the privileged MOVEC instruction at CPU space address 0xC05 and another located in the SCM at IPSBAR + 0x008. ColdFire core accesses to this memory are controlled by the processor-local copy of the RAMBAR, while module accesses are enabled by the SCM's RAMBAR.
The physical base address programmed in both copies of the RAMBAR is typically the same value; however, they can be programmed to different values. By definition, the base address must be a
0-modulo-size value.
IPSBAR
Offset:
0x008 (RAMBAR) Access: read/write
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
R
BA31 BA30 BA29 BA28 BA27 BA26 BA25 BA24 BA23 BA22 BA21 BA20 BA19 BA18 BA17 BA16
W
Reset 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
15
R 0
W
Reset 0
14
0
0
13
0
12
0
11
0
10
0
9
BDE
8
0
7
0
6
0
5
0
4
0
0 0 0 0 0 0 0 0 0 0
Figure 13-2. Memory Base Address Register (RAMBAR)
3
0
0
2
0
0
1
0
0
0
0
0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Field BA31 BA30 BA29 BA28 BA27 BA26 BA25 BA24 BA23 BA22 BA21 BA20 BA19 BA18 BA17 BA16
Reset
R/W
0000_0000_0000_0000
R/W
15 0
Field
Reset
R/W
Address
—
10 9
BDE
8
0000_0000_0000_0000
R/W
IPSBAR + 0x008
Figure 13-3.
—
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
13-5
System Control Module (SCM)
Table 13-4. RAMBAR Field Description
Field Description
31–16
BA
Base address. Defines the memory module's base address on a 64-Kbyte boundary corresponding to the physical array location within the 4 Gbyte address space supported by ColdFire.
15–10 Reserved, should be cleared.
9
BDE
Back door enable. Qualifies the module accesses to the memory.
0 Disables module accesses to the module.
1 Enables module accesses to the module.
NOTE: The SPV bit in the CPU’s RAMBAR must also be set to allow dual port access to the SRAM. For more
information, see Section 11.2.1, “SRAM Base Address Register (RAMBAR)
.”
8–0 Reserved, should be cleared.
13-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
System Control Module (SCM)
The SRAM modules are configured through the RAMBAR shown in
.
• RAMBAR specifies the base address of the SRAM.
• All undefined bits are reserved. These bits are ignored during writes to the RAMBAR and return zeros when read.
• The back door enable bit, RAMBAR[BDE], is cleared at reset, disabling the module access to the
SRAM.
NOTE
The RAMBAR default value of 0x0000_0000 is invalid. The RAMBAR located in the processor’s CPU space must be initialized with the valid bit set before the CPU (or modules) can access the on-chip SRAM (see
Chapter 11, “Static RAM (SRAM),”
for more information.
For details on the processor's view of the local SRAM memories, see Section 11.2.1, “SRAM Base
13.5.3
Core Reset Status Register (CRSR)
The CRSR contains a bit that indicates the reset source to the CPU. When the EXT bit (bit 7) reads as 1, an external device driving RSTI has caused the most recent reset. The CRSR is updated by the control logic when the reset is complete. Only one bit is set at any one time in the CRSR. The register reflects the cause of the most recent reset. To clear a bit, a logic 1 must be written to the bit location; writing a zero has no effect. Unused bits are reserved and should not be written.
NOTE
The reset status register (RSR) in the reset controller module provides indication of all reset sources except the core watchdog timer (see
Chapter 9, “Reset Controller Module” ).
IPSBAR
Offset:
0x010 (CRSR) Access: read/write
R
W
7
EXT
6
0
5
0
4
0
3
0
2
0
1
0
Reset: See Note 0 0 0 0 0 0
Note: The reset value of EXT depend on the last reset source. All other bits are initialized to zero.
Figure 13-4. Core Reset Status Register (CRSR)
0
0
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
13-7
System Control Module (SCM)
Table 13-5. CRSR Field Descriptions
Field Description
7
EXT
External reset.
1 An external device driving RSTI caused the last reset. Assertion of reset by an external device causes the processor core to initiate reset exception processing. All registers are forced to their initial state.
6–0 Reserved, should read as 0. Do not write to these locations.
13.5.4
Core Watchdog Control Register (CWCR)
The core watchdog timer prevents system lockup if the software becomes trapped in a loop with no controlled exit. The core watchdog timer can be enabled or disabled through CWCR[CWE]. It is disabled by default. If enabled, the watchdog timer requires the periodic execution of a core watchdog servicing sequence. If this periodic servicing action does not occur, the timer times out, resulting in a watchdog timer interrupt or a hardware reset, as programmed, by CWCR[CWRI]. If the timer times out and the core watchdog transfer acknowledge enable bit (CWCR[CWTA]) is set, a watchdog timer interrupt is asserted.
If a core watchdog timer interrupt acknowledge cycle has not occurred after another timeout, CWT TA is asserted in an attempt to allow the interrupt acknowledge cycle to proceed by terminating the bus cycle.
The setting of CWCR[CWTAVAL] indicates that the watchdog timer TA was asserted.
When the core watchdog timer times out and CWCR[CWRI] is programmed for a software reset, an internal reset is asserted and CRSR[CWDR] is set. To prevent the core watchdog timer from interrupting or resetting, the CWSR must be serviced by performing the following sequence:
1. Write 0x55 to CWSR.
2. Write 0xAA to the CWSR.
Both writes must occur in order before the time-out, but any number of instructions can be executed between the two writes. This order allows interrupts and exceptions to occur, if necessary, between the two writes. Caution should be exercised when changing CWCR values after the software watchdog timer has been enabled with the setting of CWCR[CWE], because it is difficult to determine the state of the core watchdog timer while it is running. The countdown value is constantly compared with the time-out period specified by CWCR[CWT]. The following steps must be taken to change CWT:
1. Disable the core watchdog timer by clearing CWCR[CWE].
2. Reset the counter by writing 0x55 and then 0xAA to CWSR.
3. Update CWCR[CWT].
4. Re-enable the core watchdog timer by setting CWCR[CWE]. This step can be performed in step 3.
The CWCR controls the software watchdog timer, time-out periods, and software watchdog timer transfer acknowledge. The register can be read at any time, but can be written only if the CWT is not pending. At system reset, the software watchdog timer is disabled.
13-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
System Control Module (SCM)
IPSBAR
Offset:
0x011 (CWCR)
7
R
W
Reset:
CWE
0
6
CWRI
5 4
CWT[2:0]
3 2
CWTA
0 0 0 0 0
Figure 13-5. Core Watchdog Control Register (CWCR)
1
CWTAVAL
0
Access: read/write
0
CWTIC
0
Table 13-6. CWCR Field Description
Field
CWT [2:0]
100
101
110
111
000
001
010
011
Description
7
CWE
6
CWRI
Core watchdog enable.
0 SWT disabled.
1 SWT enabled.
Core watchdog interrupt select.
0 If a time-out occurs, the CWT generates an interrupt to the processor core. The interrupt level for the CWT is programmed in the interrupt control register 7 (ICR7) of INTC0.
1 Reserved. If a one is written undetermined behavior will result.
Note: If a core reset is required, the watchdog interrupt should set the soft reset bit in the interrupt controller.
5–3
CWT[2:0]
Core watchdog timing delay. These bits select the timeout period for the CWT as shown in the following table. At system reset, the CWT field is cleared signaling the minimum time-out period but the watchdog is disabled
(CWCR[CWE] = 0). the following table shows the core watchdog timer delay.
C WT Time-Out Period
2
9
Bus clock frequency
2
11
Bus clock frequency
2
13
Bus clock frequency
2
15
Bus clock frequency
2
19
Bus clock frequency
2
23
Bus clock frequency
2
27
Bus clock frequency
2
31
Bus clock frequency
2
CWTA
Core watchdog transfer acknowledge enable.
0 CWTA Transfer acknowledge disabled.
1 CWTA Transfer Acknowledge enabled. After one CWT time-out period of the unacknowledged assertion of the
CWT interrupt, the transfer acknowledge asserts, which allows CWT to terminate a bus cycle and allow the interrupt acknowledge to occur.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
13-9
System Control Module (SCM)
Table 13-6. CWCR Field Description (continued)
1
CWTAVA
L
Core watchdog transfer acknowledge valid.
0 CWTA Transfer Acknowledge has not occurred.
1 CWTA Transfer Acknowledge has occurred. Write a 1 to clear this flag bit.
0
CWTIF
Core watchdog timer interrupt flag.
0 CWT interrupt has not occurred
1 CWT interrupt has occurred. Write a 1 to clear the interrupt request.
13.5.5
Core Watchdog Service Register (CWSR)
The software watchdog service sequence must be performed using the CWSR as a data register to prevent a CWT time-out. The service sequence requires two writes to this data register: first a write of 0x55 followed by a write of 0xAA. Both writes must be performed in this order prior to the CWT time-out, but any number of instructions or accesses to the CWSR can be executed between the two writes. If the CWT
illustrates the CWSR. At system reset, the contents of CWSR are uninitialized.
Access: read/write IPSBAR
Offset:
0x013 (CWSR)
7
R
W
Reset: Uninitialized
6 5 4
CWSR[7:0]
3 2
Figure 13-6. Core Watchdog Service Register (CWSR)
1 0
13-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
System Control Module (SCM)
13.6
Internal Bus Arbitration
The internal bus arbitration is performed by the on-chip bus arbiter, which containing the arbitration logic that controls which of up to four MBus masters (M0–M3 in
) has access to the external buses.
The function of the arbitration logic is described in this section.
Figure 13-7. Arbiter Module Functions
“back door” to SRAM and Flash
SRAM1
MPARK RAMBAR
CPU
M0
DMA
M2
Internal
Bus
Master
M1
MARB
Internal
Modules
13.6.1
Overview
The basic functionality is that of a 2-port, pipelined internal bus arbitration module with the following attributes:
• The master pointed to by the current arbitration pointer may get on the bus with zero latency if the address phase is available. All other requesters face at least a one cycle arbitration pipeline delay in order to meet bus timing constraints on address phase hold.
• If a requester will get an immediate address phase (that is, it is pointed to by the current arbitration pointer and the bus address phase is available), it will be the current bus master and is ignored by arbitration. All remaining requesting ports are evaluated by the arbitration algorithm to determine the next-state arbitration pointer.
• There are two arbitration algorithms: fixed and round-robin. Fixed arbitration sets the next-state arbitration pointer to the highest priority requester. Round-robin arbitration sets the next-state arbitration pointer to the highest priority requester (calculated by adding a requester's fixed priority to the current bus master’s fixed priority and then taking this sum modulo the number of possible bus masters).
• The default priority is DMA (M2) > internal master (M1) > CPU (M0), where M2 is the highest and M0 the lowest priority.
• There are two actions for an idle arbitration cycle, either leave the current arbitration pointer as is or set it to the lowest priority requester.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
13-11
System Control Module (SCM)
• The anti-lock-out logic for the fixed priority scheme forces the arbitration algorithm to round-robin if any requester has been held for longer than a specified cycle count.
13.6.2
Arbitration Algorithms
There are two modes of arbitration: fixed and round-robin. This section discusses the differences between them.
13.6.2.1
Round-Robin Mode
Round-robin arbitration is the default mode after reset. This scheme cycles through the sequence of masters as specified by MPARK[M n _PRTY] bits. Upon completion of a transfer, the master is given the lowest priority and the priority for all other masters is increased by one.
M2 =01 M1 = 10 M0 = 00 next +1 M2 =10 M1 = 11 M0 = 01 next +2 M2 =11 M1 = 00 M0 = 10 next +3 M2 =00 M1 = 01 M0 = 11
If no masters are requesting, the arbitration unit must park, pointing at one of the masters. There are two possibilities: park the arbitration unit on the last active master, or park pointing to the highest priority master. Setting MPARK[PRK_LAST] causes the arbitration pointer to be parked on the highest priority master. In round-robin mode, programming the timeout enable and lockout bits MPARK[13,11:8] will have no effect on the arbitration.
13.6.2.2
Fixed Mode
In fixed arbitration, the master with highest priority (as specified by the MPARK[M n _PRTY] bits) will win the bus. That master will relinquish the bus when all transfers to that master are complete.
If MPARK[TIMEOUT] is set, a counter will increment for each master for every cycle it is denied access.
When a counter reaches the limit set by MPARK[LCKOUT_TIME], the arbitration algorithm will be changed to round-robin arbitration mode until all locks are cleared. The arbitration will then return to fixed mode and the highest priority master will be granted the bus.
As in round-robin mode, if no masters are requesting, the arbitration pointer will park on the highest priority master if MPARK[PRK_LAST] is set, or will park on the master whose last requested the bus if cleared.
13.6.3
Bus Master Park Register (MPARK)
The MPARK controls the operation of the system bus arbitration module. The platform bus master connections are defined as the following:
• Master 2 (M2): 4-channel DMA
• Master 1 (M1): Internal Bus Master (not used in normal user operation)
• Master 0 (M0): V2 ColdFire Core
13-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
System Control Module (SCM)
IPSBAR
Offset:
0x01C (MPARK)
31
R 0
W
Reset 0
30
0
0
29
0
1
28
0
1
27
0
0
26
0
0
25 24
M2_P
_EN
BCR2
4BIT
0 0
23
0
1
22
0
1
Access: read/write
21 20
M2_PRTY
1 0
19 18
M0_PRTY
0 0
3
0
2
0
15
R 0
W
Reset 0
14
FIXED
13
TIME
OUT
12
PRKL
AST
0 0 0
11 10 9
LCKOUT_TIME
8 7
0
6
0
5
0
4
0
0 0 0 0 0 0 0 0
Figure 13-8. Default Bus Master Park Register (MPARK)
0 0
17 16
M1_PRTY
0 1
1
0
0
0
0 0
Field
31–26
25
M2_P_EN
24
BCR24BIT
23–22
21–20
M2_PRTY
19–18
M0_PRTY
17–16
15
14
FIXED
Table 13-7. MPARK Field Description
Description
Reserved, should be cleared.
DMA bandwidth control enable
0 disable the use of the DMA's bandwidth control to elevate the priority of its bus requests.
1 enable the use of the DMA's bandwidth control to elevate the priority of its bus requests.
Enables the use of 24 bit byte count registers in the DMA module
0 DMA BCRs function as 16 bit counters.
1 DMA BCRs function as 24 bit counters.
Reserved, should be cleared.
Master priority level for master 2 (DMA Controller)
00 fourth (lowest) priority
01 third priority
10 second priority
11 first (highest) priority
Master priority level for master 0 (ColdFire Core)
00 fourth (lowest) priority
01 third priority
10 second priority
11 first (highest) priority
Master priority level for master 1 (Not used in user mode)
00 fourth (lowest) priority
01 third priority
10 second priority
11 first (highest) priorityReserved, should be cleared.
Reserved, should be cleared.
Fixed or round robin arbitration
0 round robin arbitration
1 fixed arbitration
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
13-13
System Control Module (SCM)
Table 13-7. MPARK Field Description (continued)
Field Description
13
TIMEOUT
12
PRKLAST
Timeout Enable
0 disable count for when a master is locked out by other masters.
1 enable count for when a master is locked out by other masters and allow access when LCKOUT_TIME is reached.
Park on the last active master or highest priority master if no masters are active
0 park on last active master
1 park on highest priority master
11–8
LCKOUT_TIME
Lock-out Time. Lock-out time for a master being denied the bus.
The lock out time is defined as 2^ LCKOUT_TIME[3:0].
7–0 Reserved, should be cleared.
The initial state of the master priorities is M3 > M2 > M1 > M0. System software should guarantee that the programmed M n _PRTY fields are unique, otherwise the hardware defaults to the initial-state priorities.
NOTE
The M1_PRTY field should not be set for a priority higher than third
(default).
13.7
System Access Control Unit (SACU)
This section details the functionality of the system access control unit (SACU), which provides the mechanism needed to implement secure bus transactions to the address space mapped to the internal modules.
13.7.1
Overview
The SACU supports the traditional model of two privilege levels: supervisor and user. Typically, memory references with the supervisor attribute have total accessibility to all the resources in the system, while user mode references cannot access system control and configuration registers. In many systems, the operating system executes in supervisor mode, while application software executes in user mode.
The SACU further partitions the access control functions into two parts: one control register defines the privilege level associated with each bus master, and another set of control registers define the access levels associated with the peripheral modules and memory space.
The SACU’s programming model is physically implemented as part of the system control module (SCM) with the actual access control logic included as part of the arbitration controller. Each bus transaction targeted for the IPS space is first checked to see if its privilege rights allow access to the given memory space. If the privilege rights are correct, the access proceeds on the bus. If the privilege rights are insufficient for the targeted memory space, the transfer is immediately aborted and terminated with an exception, and the targeted module is not accessed.
13-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
System Control Module (SCM)
13.7.2
Features
Each bus transfer can be classified by its privilege level and the reference type. The complete set of access types includes the following:
• Supervisor instruction fetch
• Supervisor operand read
• Supervisor operand write
• User instruction fetch
• User operand read
• User operand write
Instruction fetch accesses are associated with the execute attribute.
It should be noted that while the bus does not implement the concept of reference type (code versus data) and only supports the user/supervisor privilege level, the reference type attribute is supported by the system bus. Accordingly, the access checking associated with both privilege level and reference type is performed in the IPS controller using the attributes associated with the reference from the system bus.
The SACU partitions the access control mechanisms into three distinct functions:
• Master privilege register (MPR)
— Allows each bus master to be assigned a privilege level:
– Disable the master’s user/supervisor attribute and force to user mode access
– Enable the master’s user/supervisor attribute
— The reset state provides supervisor privilege to the processor core (bus master 0).
— Input signals allow the non-core bus masters to have their user/supervisor attribute enabled at reset. This is intended to support the concept of a trusted bus master, and also controls the ability of a bus master to modify the register state of any of the SACU control registers; that is, only trusted masters can modify the control registers.
• Peripheral access control registers (PACRs)
— Provide read/write access rights, supervisor/user privilege levels.
— Reset state provides supervisor-only read/write access to these modules.
— Nine 8-bit registers control access to 17 of the on-chip peripheral modules
• Grouped peripheral access control registers (GPACR0, GPACR1)
— Provide read/write/execute access rights, supervisor/user privilege levels.
— One single register (GPACR0) controls access to 14 of the on-chip peripheral modules.
— One register (GPACR1) controls access for IPS reads and writes to the Flash module.
— Reset state provides supervisor-only read/write access to each of these peripheral spaces.
13.7.3
Memory Map/Register Definition
The memory map for the SACU program-visible registers within the system control module (SCM) is shown in
. The MPR, PACR, and GPACRs are 8 bits wide.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
13-15
System Control Module (SCM)
IPSBAR
Offset
0x020
0x024
0x028
0x02c
0x030
0x034
0x038
0x03C
[31:28] [27:24]
MPR
PACR0
PACR4
PACR8
GPACR0
—
—
—
Table 13-8. SACU Register Memory Map
[23:20] [19:16]
PPMRS
PACR1
PACR5
—
GPACR1
—
—
—
[15:12]
PPMRC
PACR2
PACR6
—
—
—
—
—
[11:8] [7:4]
IPSBMT
PACR3
PACR7
—
—
—
—
—
[3:0]
13.7.3.1
Master Privilege Register (MPR)
The MPR specifies the access privilege level associated with each bus master in the platform. The register provides one bit per bus master, where bit 3 corresponds to master 3 (fast Ethernet controller), bit 2 to master 2 (DMA Controller), bit 1 to master 1 (internal bus master), and bit 0 to master 0 (ColdFire core).
Access: read/write IPSBAR
Offset:
0x020 (MPR)
7
0 R
W
Reset: 0
6
0
5
0
4
0
3 2
MPR[3:0]
1
1
0
1 0 0 0 0
Figure 13-9. Master Privilege Register (MPR)
0
Table 13-9. MPR[n] Field Descriptions
Field Description
7–4 Reserved. Should be cleared.
3–0
MPR
Each 1-bit field defines the access privilege level of the given bus master
0 All bus master accesses are in user mode.
1 All bus master accesses use the sourced user/supervisor attribute.
n .
Only trusted bus masters can modify the access control registers. If a non-trusted bus master attempts to write any of the SACU control registers, the access is aborted with an error termination and the registers remain unaffected.
The processor core is connected to bus master 0 and is always treated as a trusted bus master. Accordingly,
MPR[0] is forced to 1 at reset.
13-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
System Control Module (SCM)
13.7.3.2
Peripheral Access Control Registers (PACR0–PACR8)
Access to several on-chip peripherals is controlled by shared peripheral access control registers. A single
PACR defines the access level for each of the two modules. These modules only support operand reads
that they control, refer to
.
IPSBAR
Offset:
0x24 + Offset (PACR n )
7 6
R
W
Reset:
LOCK1
0
5
ACCESS_CTRL1
4 3
LOCK0
2
ACCESS_CTRL0
0 0 0 0 0
Figure 13-10. Peripheral Access Control Register (PACR n )
1
0
Access: read/write
0
0
Table 13-10. PACR Field Descriptions
Field Description
7
LOCK1
6–4
ACCESS_CTRL1
This 3-bit field defines the access control for the given platform peripheral.
The encodings for this field are shown in
3
LOCK0
This bit, when set, prevents subsequent writes to ACCESSCTRL1. Any attempted write to the PACR generates an error termination and the contents of the register are not affected. Only a system reset clears this flag.
This bit, when set, prevents subsequent writes to ACCESSCTRL0. Any attempted write to the PACR generates an error termination and the contents of the register are not affected. Only a system reset clears this flag.
2–0
ACCESS_CTRL0
This 3-bit field defines the access control for the given platform peripheral.
The encodings for this field are shown in
Table 13-11. PACR ACCESSCTRL Bit Encodings
Bits
100
101
110
111
000
001
010
011
Supervisor Mode
Read/Write
Read
Read
Read
Read/Write
Read/Write
Read/Write
No Access
User Mode
No Access
No Access
Read
No Access
Read/Write
Read
Read/Write
No Access
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
13-17
System Control Module (SCM)
Table 13-12. Peripheral Access Control Registers (PACRs)
Modules Controlled
IPSBAR Offset Name
0x024
0x025
0x026
0x027
0x028
0x029
0x02a
0x02b
0x02c
0x02d
0x02e
PACR0
PACR1
PACR2
PACR3
PACR4
—
PACR6
PACR7
PACR8
—
—
ACCESS_CTRL1
SCM
—
UART0
UART2
I
2
C
—
DTIM0
DTIM2
INTC0
—
—
ACCESS_CTRL0
—
DMA
UART1
—
QSPI
—
DTIM1
DTIM3
—
—
—
At reset, these on-chip modules are configured to have only supervisor read/write access capabilities. If an instruction fetch access to any of these peripheral modules is attempted, the IPS bus cycle is immediately terminated with an error.
13.7.3.3
Grouped Peripheral Access Control Registers (GPACR0 & GPACR1)
The on-chip peripheral space starting at IPSBAR is subdivided into 16 64-Mbyte regions. Each of the first two regions has a unique access control register associated with it. The other 14 regions are in reserved space; the access control registers for these regions are not implemented. Bits [29:26] of the address select the specific GPACRn to be used for a given reference within the IPS address space. These access control registers are 8 bits wide so that read, write, and execute attributes may be assigned to the given IPS region.
NOTE
The access control for modules with memory space protected by
PACR0–PACR8 are determined by the PACR0–PACR8 settings. The access control is not affected by GPACR0, even though the modules are mapped in its 64-Mbyte address space.
Access: read/write IPSBAR
Offset:
0x030, IPSBAR + 0x31 (GPACR)
7 6
0 R
W
Reset:
LOCK
0 0
5
0
4
0
3
0 0 0
Figure 13-11. GPACR Register
2 1
ACCESS_CTRL
0 0
0
0
13-18
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
System Control Module (SCM)
Table 13-13. Grouped Peripheral Access Control Register (GPACR) Field Descriptions
Field Description
7
LOCK
6–4
This bit, once set, prevents subsequent writes to the GPACR. Any attempted write to the GPACR generates an error termination and the contents of the register are not affected. Only a system reset clears this flag.
Reserved, should be cleared.
3–0
ACCESS_CTRL
This 4-bit field defines the access control for the given memory region.
The encodings for this field are shown in Table 13-14
.
At reset, these on-chip modules are configured to have only supervisor read/write access capabilities. Bit
encodings for the ACCESS_CTRL field in the GPACR are shown in Table 13-14 .
shows the memory space protected by the GPACRs and the modules mapped to these spaces.
Table 13-14. GPACR ACCESS_CTRL Bit Encodings
Bits
1000
1001
1010
1011
1100
1101
1110
1111
0000
0001
0010
0011
0100
0101
0110
0111
Supervisor Mode
Read / Write
Read
Read
Read
Read / Write
Read / Write
Read / Write
No Access
Read / Write / Execute
Read / Execute
Read / Execute
Execute
Read / Write / Execute
Read / Write / Execute
Read / Write
Read / Write / Execute
User Mode
No Access
No Access
Read
No Access
Read / Write
Read
Read / Write
No Access
No Access
No Access
Read / Execute
No Access
Read / Write / Execute
Read / Execute
Read
Execute
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
13-19
System Control Module (SCM)
Register
GPACR0
GPACR1
Table 13-15. GPACR Address Space
Space Protected
(IPSBAR Offset)
0x0000_0000–
0x03FF_FFFF
Modules Protected
0x0400_0000–
0x07FF_FFFF
Ports, CCM, PMM, Reset controller, Clock,
EPORT, WDOG, PIT0–PIT3, QADC, GPTA,
GPTB, FlexCAN, CFM (Control)
CFM (Flash module’s backdoor access for programming or access by a bus master other than the core)
13-20
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 14
General Purpose I/O Module
14.1
Introduction
Many of the pins associated with the external interface may be used for several different functions. When not used for their primary function, many of the pins may be used as general-purpose digital I/O pins. In some cases, the pin function is set by the operating mode, and the alternate pin functions are not supported.
The digital I/O pins are grouped into 8-bit ports. Some ports do not use all 8 bits. Each port has registers that configure, monitor, and control the port pins.
is a block diagram of the MCF52235 ports.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
14-1
General Purpose I/O Module
PORT DD
PORT AS
PORT QS
PORT TA
DDATA[3:0] / PDD[7:4]
PST[3:0] / PDD[3:0]
CANTX / PAS[3] / SYNCB
CANRX / PAS[2] / SYNCA
SDA / PAS[1] / CANRX / RXD2
SCL / PAS[0] / CANTX / TXD2
QSPI_PCS3 / PQS[6] / SYNCA / SYNCB
QSPI_PCS2 / PQS[5]
QSPI_PCS1 / PQS[4]
QSPI_PCS0 / PQS[3] / SDA / CTS1
QSPI_SCK / PQS[2] / SCL / RTS1
QSPI_DIN / PQS[1] / CANRX / RXD1
QSPI_DOUT / PQS[0] / CANTX TXD1
GPT[3] / PTA[3] / PWM7
GPT[2] / PTA[2] / PWM5
GPT[1] / PTA[1] / PWM3
GPT[0] / PTA[0] / PWM1
PORT TD
PORT TC
PWM7 / PTD[3]
PWM5 / PTD[2]
PWM3 / PTD[1]
PWM1 / PTD[0]
DTIN3 / PTC[3] / DTOUT3 / PWM6
DTIN2 / PTC[2] / DTOUT2 / PWM4
DTIN1 / PTC[1] / DTOUT1 / PWM2
DTIN0 / PTC[0] / DTOUT0 / PWM0
PORT UC
PORT UB
PORT UA
PORT AN
PORT NQ
PORT LD
PORT GP
CTS2 / PUC[3]
RTS2 / PUC[2]
RXD2 / PUC[1]
TXD2 / PUC[0]
CTS1 / PUB[3] / SYNCA / RXD2
RTS1 / PUB[2] / SYNCB / TXD2
RXD1 / PUB[1]
TXD1 / PUB[0]
Figure 14-1. General Purpose I/O Module Block Diagram
14.2
Overview
The MCF52235 ports module controls the configuration for the following external pins:
• External bus accesses
• Chip selects
• Debug data
• Processor status
• FlexCAN transmit/receive data
• I
2
C serial control
• QSPI
• UART transmit/receive
• 32-bit DMA timers
CTS0 / PUA[3] / CANRX
RTS0 / PUA[2] / CANTX
RXD0 / PUA[1]
TXD0 / PUA[0]
AN0 / PAN[0]
AN1 / PAN[1]
AN2 / PAN[2]
AN3 / PAN[3]
AN4 / PAN[4]
AN5 / PAN[5]
AN6 / PAN[6]
AN7 / PAN[7]
IRQ1 / PNQ[1] / SYNCA / PWM1
IRQ2 / PNQ[2]
IRQ3 / PNQ[3]
IRQ4 / PNQ[4]
IRQ5 / PNQ[5]
IRQ6 / PNQ[6]
IRQ7 / PNQ[7]
ACTLED / PLD[0]
LNKLED / PLD[1]
SPDLED / PLD[2]
DUPLED / PLD[3]
COLLED / PLD[4]
RXLED / PLD[5]
TXLED / PLD[6]
IRQ8 / PGP[0]
IRQ9 / PGP[1]
IRQ10 / PGP[2]
IRQ11 / PGP[3]
IRQ12 / PGP[4]
IRQ13 / PGP[5]
14-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
General Purpose I/O Module
14.3
Features
The MCF52235 ports includes these distinctive features:
• Control of primary function use on all ports
• Digital I/O support for all ports; registers for:
— Storing output pin data
— Controlling pin data direction
— Reading current pin state
— Setting and clearing output pin data registers
14.4
Signal Descriptions
Refer to Chapter 2, “Signal Descriptions ,” for more detailed information on the different signals and pins.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
14-3
General Purpose I/O Module
14.5
Memory Map/Register Definition
14.5.1
Ports Memory Map
14-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
General Purpose I/O Module
Table 14-1. Ports Module Memory Map
Offset 1 31–24 23–16 15–8 7–0 Access 2
Port Output Data Registers
$0000
$0004
$0008
$000C
$0010
$0014
PORTNQ
PORTQS
PORTTD
PORTDD
Port Data Direction Registers
$0018
$001C
$0020
$0024
$0028
$002C
DDRNQ
DDRQS
DDRTD
DDRDD
Port Pin Data/Set Data Registers
$0030
$0034
$0038
$003C
$0040
$0044
PORTNQP/SETNQ
PORTQSP/SETQS
PORTTDP/SETTD
PORTDDP/SETDD
Port Clear Output Data Registers
$0048
$004C
$0050
$0054
$0058
$005C
CLRNQ
CLRQS
CLRTD
CLRDD
Reserved
Reserved
PORTUA
PORTLD
Reserved
Reserved
DDRUA
DDRLD
Reserved
Reserved
Reserved
Reserved
CLRUA
CLRLD
Reserved
Reserved
Reserved
Reserved
PORTUAP/SETUA
PORTLDP/SETLD
Reserved
Reserved
PORTAN
PORTTA
PORTUB
PORTGP
DDRAN
DDRTA
DDRUB
DDRGP
PORTANP/SETAN
PORTTAP/SETTA
PORTUBP/SETUB
PORTGPP/SETGP
Reserved
Reserved
CLRAN
CLRTA
CLRUB
CLRGP
PORTAS
PORTTC
PORTUC
Reserved
DDRAS
DDRTC
DDRUC
Reserved
PORTASP/SETAS
PORTTCP/SETTC
PORTUCP/SETUC
Reserved
CLRAS
CLRTC
CLRUC
Reserved
S/U
S/U
S/U
S/U
S/U
S/U
S/U
S/U
S/U
S/U
S/U
S/U
S/U
S/U
S/U
Port Pin Assignment Registers
$0060
$0064
$0068 PNQPAR
PQSPAR $006C
$0070
$0074
PTDPAR
PDDPAR
Port Pad Control Registers
PUAPAR
PLDPAR
Reserved
Reserved
PANPAR
PTAPAR
PUBPAR
PGPPAR
PASPAR
PTCPAR
PUCPAR
Reserved
S/U
S/U
S/U
S/U
S/U
S/U
$0078
$007C
PWOR[15:0]
PDSR[31:0]
PDRR[37:32] S/U
S/U
1
The register address is the sum of the IPSBAR address and the base address offset.
2
S/U = supervisor or user mode access. User mode accesses to supervisor-only addresses have no effect and cause a cycle termination transfer error.
S/U
S/U
S/U
S/U
S/U
S/U
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
14-5
General Purpose I/O Module
14.6
Register Descriptions
14.6.1
Port Output Data Registers (PORT
n
)
The PORT n registers store the data to be driven on the corresponding port n pins when the pins are configured for digital output.
The PORT n registers with a full 8-bit implementation are shown in Figure 11-2 . The remaining PORT n registers use fewer than 8 bits. Their bit definitions are shown in Figure 11-3 , Figure 11-4 , Figure 11-5 , and Figure 11-6 .
The PORT n registers are read/write. At reset, all bits in the PORT n registers are set.
Reading a PORT n register returns the current values in the register, not the port n pin values.
PORT n bits can be set by setting the PORT n register, or by setting the corresponding bits in the
PORT n P/SET n register. They can be cleared by clearing the PORT n register, or by clearing the corresponding bits in the CLR n register.
IPSBAR
Offset:
Base + $000A (PORTAN)
Base + $0014 (PORTDD)
Base + $0016 (PORTGP)
7 6
R
W
Reset:
PORT
1 n 7 PORT n 6
5
PORT n 5
4
PORT n 4
3
PORT n 3
2
PORT n 2
1 1 1 1 1
Figure 14-2. — Port Output Data Registers [7:0]
Access: User read/write
1
PORT n 1 PORT n 0
1
0
1
IPSBAR
Offset:
Base + $000B (PORTAS)
Base + $000E (PORTTA)
Base + $000F (PORTTC)
Base + $0010 (PORTTD)
Base + $0011 (PORTUA)
Base + $0012 (PORTUB)
Base + $0013 (PORTUC)
7
0
6
0 R
W
Reset: 0
5
0
4
0
3
PORT n 3
2
PORT n 2
0 0 0 1 1
Figure 14-3. — Port Output Data Registers [3:0]
Access: User read/write
1 0
PORT n 1 PORT n 0
1 1
14-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
General Purpose I/O Module
IPSBAR
Offset:
Base + $000C (PORTQS)
Base + $0015 (PORTLD)
6 7
0
R
W
Reset: 0
5
PORT n 6 PORT n 5
4
PORT n 4
3
PORT n 3
2
PORT n 2
1 1 1 1 1
Figure 14-4. — Port Output Data Registers [6:0]
Access: User read/write
1
PORT n 1 PORT n 0
1
0
1
IPSBAR
Offset:
Base + $0008 (PORTNQ)
7 6 5 4 3 2
R
W
Reset:
PORT n 7 PORT n 6 PORT n 5 PORT n 4 PORT n 3 PORT n 2
1 1 1 1 1 1
Figure 14-5. — Port Output Data Registers [7:1]
PORTn
Port n output data bits.
Drive 1 when port n pin is digital output.
Drive 0 when port n pin is digital output.
Access: User read/write
1
PORT n 1
1
0
0
0
14.6.2
Port Data Direction Registers (DDR
n)
The DDR n registers control the direction of the port n pin drivers when the pins are configured for digital
I/O.
The DDR n registers are read/write. At reset, all bits in the DDR n registers are cleared to 0s.
Setting any bit in a DDR n register configures the corresponding port n pin as an output. Clearing any bit in a DDR n register configures the corresponding pin as an input.
IPSBAR
Offset:
Base + $0022 (DDRAN)
Base + $002C (DDRDD)
Base + $002E (DDRGP)
7 6
R
W
Reset:
DDR
0 n 7 DDR n 6
5
DDR n 5
4
DDR n 4
3
DDR n 3
2
DDR n 2
0 0 0 0 0
Figure 14-6. — Port Data Direction Registers [7:0]
Access: User read/write
1
DDR n 1
0
0
DDR n 0
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
14-7
General Purpose I/O Module
IPSBAR
Offset:
Base + $0023 (DDRAS)
Base + $0026 (DDRTA)
Base + $0027 (DDRTC)
Base + $0028 (DDRTD)
Base + $0029 (DDRUA)
Base + $002A (DDRUB)
Base + $002B (DDRUC)
7
0
6
0 R
W
Reset: 0
5
0
4
0
3
DDR n 3
2
DDR n 2
0 0 0 0 0
Figure 14-7. — Port Data Direction Registers [3:0]
Access: User read/write
1
DDR n 1
0
0
DDR n 0
0
IPSBAR
Offset:
Base + $0024 (DDRQS)
Base + $002D (DDRLD)
R
W
Reset:
7
0
6
DDR n 6
5
DDR n 5
4
DDR n 4
3
DDR n 3
2
DDR n 2
0 0 0 0 0 0
Figure 14-8. — Port Data Direction Registers [6:0]
DDR n
IPSBAR
Offset:
Base + $001C (DDRNQ)
7 6 5 4 3 2
R
W
Reset:
DDR n 7 DDR n 6 DDR n 5 DDR n 4 DDR n 3 DDR n 2
0 0 0 0 0 0
Figure 14-9. — Port Data Direction Registers [7:1]
Port n data direction bits.
1 = Port n pin configured as output.
0 = Port n pin configured as input.
Access: User read/write
1
DDR n 1
0
1
DDR n 1
0
DDR n 0
0
Access: User read/write
0
0
0
0
14.6.3
Port Pin Data/Set Data Registers (PORT
n
P/SET
n
)
The PORT n P/SET n registers reflect the current pin states and control the setting of output pins when the pin is configured for digital I/O.
The PORT n P/SET n registers are read/write. At reset, the bits in the PORT n P/SET n registers are set to the current pin states.
Reading a PORT n P/SET n register returns the current state of the port n pins.
14-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
General Purpose I/O Module
Writing 1s to a PORT n P/SET n register sets the corresponding bits in the PORT n register. Writing 0s has no effect.
IPSBAR
Offset:
Base + $003A (PORTAN/SETAN)
Base + $0044 (PORTDD/SETDD)
Base + $0046 (PORTGPP/SETGP)
Access: User read/write
7 6 5 4 3 2 1 0
R
W
Reset:
PORT n P7 PORT n P6 PORT n P5 PORT n P4 PORT n P3 PORT n P2 PORT n P1 PORT n P0
1 1 1 1 1 1 1 1
Figure 14-10. PORT n P/SET n — Port Pin Data/Set Data Registers [7:0]
IPSBAR
Offset:
Base + $003B (PORTAS/SETAS)
Base + $003E (PORTTA/SETTA)
Base + $003F (PORTTC/SETTC)
Base + $0040 (PORTTD/SETTD)
Base + $0041 (PORTUA/SETUA)
Base + $0042 (PORTUB/SETUB)
Base + $0043 (PORTUC/SETUC)
7
0
6
0
R
W
Reset:
5
0
4
0
3 2
Access: User read/write
1 0
PORT n P3 PORT n P2 PORT n P1 PORT n P0
0 0 0 0 1 1 1
Figure 14-11. PORT n P/SET n — Port Pin Data/Set Data Registers [3:0]
1
IPSBAR
Offset:
Base + $003C (PORTQS/SETQS)
Base + $0045 (PORTLD/SETLD)
6 5 7
0
R
W
Reset:
4 3 2
Access: User read/write
1 0
PORT n P6 PORT n P5 PORT n P4 PORT n P3 PORT n P2 PORT n P1 PORT n P0
1 0 1 1 1 1 1 1
Figure 14-12. PORT n P/SET n — Port Pin Data/Set Data Registers [6:0]
IPSBAR
Offset:
Base + $0038 (PORTNQ/SETNQ) Access: User read/write
7 6 5 4 3 2 1
R
W
Reset:
PORT n P7 PORT n P6 PORT n P5 PORT n P4 PORT n P3 PORT n P2 PORT n P1
1 1 1 1 1 1 1
Figure 14-13. /SET n — Port Pin Data/Set Data Registers [7:1]
0
0
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
14-9
General Purpose I/O Module
PORT n P/SET n
Port n pin data/set data bits.
1 = Port n pin state is 1 (read); set corresponding PORT n bit (write).
0 = Port n pin state is 0 (read).
14.6.4
Port Clear Output Data Registers (CLR
n
)
Writing 0s to a CLR n register clears the corresponding bits in the PORT n register. Writing 1s has no effect.
Reading the CLR n register returns 0s.
The CLR n registers are read/write.
IPSBAR
Offset:
Base + $005C (CLRDD)
Base + $0052 (CLRAN)
Base + $0054 (CLRGP)
7 6
R
W
Reset:
CLR
0 n 7 CLR n 6
5
CLR n 5
4
CLR n 4
3
CLR n 3
2
CLR n 2
0 0 0 0 0
Figure 14-14. — Port Clear Output Data Registers [7:0]
Access: User read/write
1
CLR n 1
0
0
CLR n 0
0
IPSBAR
Offset:
Base + $0053 (CLRAS)
Base + $0056 (CLRTA)
Base + $0057 (CLRTC)
Base + $0058 (CLRTD)
Base + $0059 (CLRUA)
Base + $005A (CLRUB)
Base + $005B (CLRUC)
7
0
6
0
R
W
Reset: 0
5
0
4
0
3
CLR n 3
2
CLR n 2
0 0 0 0 0
Figure 14-15. — Port Clear Output Data Registers [3:0]
Access: User read/write
1
CLR n 1
0
0
CLR n 0
0
IPSBAR
Offset:
Base + $0054 (CLRQS)
Base + $005D (CLRLD)
6 7
0
R
W
Reset: 0
CLR n 6
5
CLR n 5
4
CLR n 4
3
CLR n 3
2
CLR n 2
0 0 0 0 0
Figure 14-16. — Port Clear Output Data Registers [6:0]
Access: User read/write
1
CLR n 1
0
0
CLR n 0
0
14-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
General Purpose I/O Module
IPSBAR
Offset:
Base + $0050 (CLRNQ) Access: User read/write
7 6 5 4 3 2
R
W
Reset:
CLR n 7 CLR n 6 CLR n 5 CLR n 4 CLR n 3 CLR n 2
0 0 0 0 0 0
Figure 14-17. — Port Clear Output Data Registers [7:1]
CLR n
Port n clear output data register bits.
1 = Never returned for reads; no effect for writes.
0 = Always returned for reads; clears corresponding PORT n bit for writes.
1
CLR n 1
0
0
0
0
14.6.5
Pin Assignment Registers
All pin assignment registers are read/write.
If multiple pins are configured for the one function, then the result is undefined.
14.6.5.1
Dual Function Pin Assignment Registers
The dual function pin assignment registers allow each pin controlled by each register bit to be configured between the GPIO function and the primary function
Access: User read/write IPSBAR
Offset:
Base + $006A (PANPAR)
Base + $0074 (PDDPAR)
Base + $0076 (PGPPAR)
7 6 5 4 3 2 1 0
R
W
Reset:
P n PAR7 P n PAR6 P n PAR5 P n PAR4 P n PAR3 P n PAR2 P n PAR1 P n PAR0
0 0 0 0 0 0 0
Figure 14-18. P n PAR — Port n Pin Assignment Registers Dual [7:0]
0
IPSBAR
Offset:
Base + $0075 (PLDPAR)
6 7
0 R
W
Reset: 0
P n PAR6
5
P n PAR5
4
P n PAR4
3
P n PAR3
2
P n PAR2
0 0 0 0 0
Figure 14-19. — Port n Pin Assignment Registers Dual [6:0]
Access: User read/write
1
P n PAR1
0
0
P n PAR0
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
14-11
General Purpose I/O Module
IPSBAR
Offset:
Base + $0070 (PTDPAR)
Base + $0073 (PUCPAR)
Access: User read/write
R
W
Reset:
7
0
6
0
5
0
4
0
3
P n PAR3
2
P n PAR2
1
P n PAR1
0 0 0 0 0 0
Figure 14-20. P n PAR — Port n Pin Assignment Registers Dual [3:0]
0
P n PAR
Port n pin assignment register bits.
1 = Pin assumes the primary function for that pin.
0 = Pin assumes the GPIO function for that pin.
0
P n PAR0
0
14.6.5.2
Quad Function Pin Assignment Registers
The quad function pin assignment registers allow each pin controlled by each register bit to be configured between the GPIO function, primary function, alternate 1 function, and the alternate 2 function
Access: User read/write IPSBAR
Offset:
Base + $0068 (PNQPAR)
15 14
R
W
Reset 0
P n PAR7
0
13
0
P n PAR6
12
0
11
0
P n PAR5
10
0
9
0
P n PAR4
8
0
R
W
Reset
7 6 5 4 3 2
P n PAR3 P n PAR2 P n PAR1
0 0 0 0 0 0
Figure 14-21. P n PAR — Port n Pin Assignment Registers Quad [15:2]
0
1
0
0
0
0
IPSBAR
Offset:
Base + $006C (PQSPAR)
15
0
14
0 R
W
Reset 0 0
R
W
Reset
7
0
13
P n PAR6
12
0 0
11
P n PAR5
10
0 0
Access: User read/write
9
P n PAR4
0
8
0
P n PAR3
6 5
P n PAR2
4 3
P n PAR1
2
0 0 0 0 0
Figure 14-22. P n PAR — Port n Pin Assignment Registers Quad [13:0]
0
1
P n PAR0
0
0
14-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
General Purpose I/O Module
IPSBAR
Offset:
Base + $006B (PASPAR)
Base + $006E (PTAPAR)
Base + $006F (PTCPAR)
Base + $0071 (PUAPAR)
Base + $0072 (PUBPAR)
Access: User read/write
7 6 5 4 3 2 1 0
R
W
Reset:
P n PAR3 P n PAR2 P n PAR1
0 0 0 0 0 0
Figure 14-23. P n PAR — Port n Pin Assignment Registers Quad [7:0]
0
P n PAR
Port n pin assignment register bits.
Table 14-2. Double Bit Pin Assignment Register Bit
Bit Value
00
01
10
11
Pin Assignment
GPIO function
Primary function
Alternate 1 function
Alternate 2 function
P n PAR0
0
14.6.5.3
Pin Wired OR Register
which Pin Wired OR register bit controls which pin.
Access: User read/write IPSBAR
Offset:
Base + $0078 (PWOR)
14
R
W
Reset
15
PWOR15
0
PWOR14
0
13
PWOR13
0
12 11 10
PWOR 12 PWOR11 PWOR 10
0 0 0
9
PWOR9
0
8
PWOR 8
0
R
W
Reset
7
PWOR7
6
PWOR
6
5
PWOR5
4
PWOR
4
3
PWOR3
2
PWOR
2
1
PWOR1
0 0 0 0 0 0
Figure 14-24. P n PAR — Port n Pin Assignment Registers Quad [15:2]
0
A logic 1 configures the selected bit for Wired OR operation
A logic 0 configures the selected bit for normal operation.
0
PWOR
0
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
14-13
General Purpose I/O Module
14.6.5.4
Pin Drive Strength Register
The pin drive strength register is read/write, and each bit resets to logic 0 in single chip mode (MCF52235 default) and logic 1 in EzPort and FAST mode.
IPSBAR
Offset:
0x00_007A (PDSR)
R
W
Reset
47
PDSR47
1)
46
PDSR46
45
PDSR45
44
PDSR44
43
PDSR43
42
PDSR42
Access: User read/write
41
PDSR41
40
PDSR40
33
PDSR33
32
PDSR32
R
W
Reset
39
PDSR39
R
W
Reset
31
PDSR31
1)
R
W
Reset
23
PDSR23
1)
R
W
Reset
15
PDSR15
1)
38
PDSR38
30
PDSR30
22
PDSR22
14
PDSR14
37
PDSR37
29
PDSR29
21
PDSR21
13
PDSR13
36
PDSR36
28
PDSR28
20
PDSR20
12
PDSR12
35
PDSR35
27
PDSR27
19
PDSR19
11
PDSR11
34
PDSR34
26
PDSR26
18
PDSR18
10
PDSR10
7 6 5 4 3
R
W
Reset
PDSR7
1)
PDSR6 PDSR5 PDSR4 PDSR3
1)Each bit resets to logic 0 in Single Chip mode and logic 1 in EzPort/FAST mode.
2
PDSR2
Figure 14-25. PDSR — Pin Drive Strength Register [31:0]
25
PDSR25
17
PDSR17
9
PDSR9
1
PDSR1
24
PDSR24
16
PDSR16
8
PDSR8
0
PDSR0
14-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
General Purpose I/O Module
.
Base + $007A (PDSR) Upper 16bits
Field
Reset
R/W
47
PDSR47
46
PDSR46
45
PDSR45
44
PDSR44
1
43
PDSR43
42
PDSR42
41
PDSR41
40
PDSR40
39 38 37 36
R/W
35 34
Field
Reset
R/W R/W
1
Each bit resets to logic 0 in Single Chip mode and logic 1 in EzPort/FAST mode.
33 32
PDSR39 PDSR38 PDSR37 PDSR36 PDSR35 PDSR34 PDSR33 PDSR32
1
Figure 14-26. PDSR — Pin Drive Strength Register [47:32]
.
Base + $007C (PDSR) Low 32 bits
Field
Reset
R/W
31
PDSR31
30
PDSR30
29
PDSR29
28
PDSR28
1
27
PDSR27
26
PDSR26
25
PDSR25
24
PDSR24
Field
Reset
R/W
23
PDSR23
22
PDSR22
21
PDSR21
20
PDSR20
R/W
1
19
PDSR19
18
PDSR18
Field
Reset
R/W
15
PDSR15
14
PDSR14
13
PDSR13
12
PDSR12
R/W
1
11
PDSR11
10
PDSR10
Field
Reset
R/W
7
PDSR7
6
PDSR6
5
PDSR5
4
PDSR4
R/W
1
R/W
3
PDSR3
2
PDSR2
1
Each bit resets to logic 0 in Single Chip mode and logic 1 in EzPort/FAST mode.
17
PDSR17
9
PDSR9
1
PDSR1
16
PDSR16
8
PDSR8
0
PDSR0
Figure 14-27. PDSR — Pin Drive Strength Register [31:0]
PDSR
Pin drive strength register control bits.
1 = Pin is configured for high drive strength (10mA).
0 = Pin is configured for low drive strength (2mA).
14.7
Ports Interrupts
The ports module does not generate interrupt requests.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
14-15
General Purpose I/O Module
14-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 15
Interrupt Controller Module
This section details the functionality for the MCF52235 interrupt controllers (INTC0, INTC1). The general features of each of the interrupt controller include the following:
• 63 interrupt sources per controller
— fully-programmable interrupt sources
— 7 fixed-level interrupt sources
• Each of the 63 sources has a unique interrupt control register (ICR nx ) to define the software-assigned levels and priorities within the level
• Unique vector number for each interrupt source
• Ability to mask any individual interrupt source, plus global mask-all capability
• Supports both hardware and software interrupt acknowledge cycles
• Wake-up signal from low-power stop modes
The 56 fully-programmable and 7 fixed-level interrupt sources for each interrupt controller handle the complete set of interrupt sources from all of the modules on the device. This section describes how the interrupt sources are mapped to the interrupt controller logic and how interrupts are serviced.
15.1
68K/ColdFire Interrupt Architecture Overview
Before continuing with the specifics of the interrupt controllers, a brief review of the interrupt architecture of the 68K/ColdFire family is appropriate.
The interrupt architecture of ColdFire is exactly the same as the M68000 family, where there is a 3-bit encoded interrupt priority level sent from the interrupt controller to the core, providing 7 levels of interrupt requests. Level 7 represents the highest priority interrupt level, while level 1 is the lowest priority. The processor samples for active interrupt requests once per instruction by comparing the encoded priority level against a 3-bit interrupt mask value (I) contained in bits 10:8 of the machine’s status register (SR). If the priority level is greater than the SR[I] field at the sample point, the processor suspends normal instruction execution and initiates interrupt exception processing. Level 7 interrupts are treated as non-maskable and edge-sensitive within the processor, while levels 1–6 are treated as level-sensitive and may be masked depending on the value of the SR[I] field. For correct operation, ColdFire requires that the interrupt source, once asserted, remains asserted until explicitly disabled by the interrupt service routine.
During the interrupt exception processing, the CPU enters supervisor mode, disables trace mode, and then fetches an 8-bit vector from the interrupt controller. This byte-sized operand fetch is known as the interrupt acknowledge (IACK) cycle, with the ColdFire implementation using a special encoding of the transfer type and transfer modifier attributes to distinguish this data fetch from a normal memory access. The fetched data provides an index into the exception vector table, which contains 256 addresses, each pointing
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
15-1
Interrupt Controller Module to the beginning of a specific exception service routine. In particular, vectors 64–255 of the exception vector table are reserved for user interrupt service routines. The first 64 exception vectors are reserved for the processor to handle reset, error conditions (access, address), arithmetic faults, system calls, etc. Once the interrupt vector number has been retrieved, the processor continues by creating a stack frame in memory. For ColdFire, all exception stack frames are 2 longwords in length and contain 32 bits of vector and status register data, along with the 32-bit program counter value of the instruction that was interrupted
(see Section 3.5, “Exception Stack Frame Definition,”
for more information on the stack frame format).
After the exception stack frame is stored in memory, the processor accesses the 32-bit pointer from the exception vector table using the vector number as the offset, and then jumps to that address to begin execution of the service routine. After the status register is stored in the exception stack frame, the SR[I] mask field is set to the level of the interrupt being acknowledged, effectively masking that level and all lower values while in the service routine. For many peripheral devices, the processing of the IACK cycle directly negates the interrupt request, while other devices require that request to be explicitly negated during the processing of the service routine.
For this device, the processing of the interrupt acknowledge cycle is fundamentally different than previous
68K/ColdFire cores. In the new approach, all IACK cycles are directly handled by the interrupt controller, so the requesting peripheral device is not accessed during IACK. As a result, the interrupt request must be explicitly cleared in the peripheral during the interrupt service routine. For more information, see
Section 15.1.1.3, “Interrupt Vector Determination .”
Unlike the M68000 family, all ColdFire processors guarantee that the first instruction of the service routine is executed before sampling for interrupts is resumed. By making this initial instruction a load of the SR, interrupts can be safely disabled if required.
During the execution of the service routine, the appropriate actions must be performed on the peripheral to negate the interrupt request.
For more information on exception processing, see the ColdFire Programmer’s Reference Manual at http://www.freescale.com/coldfire .
15.1.1
Interrupt Controller Theory of Operation
To support the interrupt architecture of the 68K/ColdFire programming model, the combined 63 interrupt sources are organized as 7 levels, with each level supporting up to 9 prioritized requests. Consider the
.
Table 15-1. Interrupt Priority Within a Level
ICR[2:0]
111
110
101
100
—
Priority
7 (Highest)
6
5
4
Fixed Midpoint Priority
Interrupt
Sources
8–63
8–63
8–63
8–63
1–7
15-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Interrupt Controller Module
Table 15-1. Interrupt Priority Within a Level (continued)
ICR[2:0]
011
010
001
000
Priority
3
2
1
0 (Lowest)
Interrupt
Sources
8–63
8–63
8–63
8–63
The level and priority is fully programmable for all sources except interrupt sources 1–7. Interrupt source
1–7 (from the edgeport module) are fixed at the corresponding level’s midpoint priority. Thus, a maximum of 8 fully-programmable interrupt sources are mapped into a single interrupt level. The fixed interrupt source is hardwired to the given level and represents the mid-point of the priority within the level. For the fully-programmable interrupt sources, the 3-bit level and the 3-bit priority within the level are defined in the 8-bit interrupt control register (ICR nx ).
The operation of the interrupt controller can be broadly partitioned into three activities:
• Recognition
• Prioritization
• Vector determination during IACK
15.1.1.1
Interrupt Recognition
The interrupt controller continuously examines the request sources and the interrupt mask register to determine if there are active requests. This is the recognition phase.
15.1.1.2
Interrupt Prioritization
As an active request is detected, it is translated into the programmed interrupt level, and the resulting 7-bit decoded priority level (IRQ[7:1]) is driven out of the interrupt controller.
15.1.1.3
Interrupt Vector Determination
Once the core has sampled for pending interrupts and begun interrupt exception processing, it generates an interrupt acknowledge (IACK) cycle. The IACK transfer is treated as a memory-mapped byte read by the processor and routed to the appropriate interrupt controller. Next, the interrupt controller extracts the level being acknowledged from address bits 4:2, determines the highest priority interrupt request active for that level, and returns the 8-bit interrupt vector for that request to complete the cycle. The 8-bit interrupt vector is formed using the following algorithm:
For INTC0, vector_number = 64 + interrupt source number
Recall vector_numbers 0–63 are reserved for the ColdFire processor and its internal exceptions. Thus, the following mapping of bit positions to vector numbers applies for the INTC0: if interrupt source 1 is active and acknowledged, if interrupt source 2 is active and acknowledged,
...
then vector_number = 65 then vector_number = 66
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
15-3
Interrupt Controller Module if interrupt source 8 is active and acknowledged, if interrupt source 9 is active and acknowledged,
...
if interrupt source 62 is active and acknowledged, then vector_number = 72 then vector_number = 73 then vector_number = 126
The net effect is a fixed mapping between the bit position within the source to the actual interrupt vector number.
If there is no active interrupt source for the given level, a special spurious interrupt vector
(vector_number = 24) is returned. It is the responsibility of the service routine to handle this error situation.
Note this protocol implies the interrupting peripheral is not accessed during the acknowledge cycle since the interrupt controller completely services the acknowledge. This means the interrupt source must be explicitly disabled in the interrupt service routine. This design provides unique vector capability for all interrupt requests, regardless of the complexity of the peripheral device.
15.2
Memory Map
The register programming model for the interrupt controllers is memory-mapped to a 256-byte space. In the following discussion, there are a number of program-visible registers greater than 32 bits. For these control fields, the physical register is partitioned into two 32-bit values: a register high (the upper longword, represented by an appended “H”) and a register low (the lower longword, represented by an appended “L”).
The registers and their locations are defined in
Table 15-3 . The offsets listed start from the base address
for each interrupt controller. The base addresses for the interrupt controllers are listed below.
Table 15-2. Interrupt Controller Base Addresses
Interrupt Controller Number
INTC
INTC1
Base Address
IPSBAR + 0xC00
IPSBAR + 0xD00
Module Offset
IPSBAR + 0x0C00
IPSBAR + 0x0C04
IPSBAR + 0x0C08
IPSBAR + 0x0C0c
IPSBAR + 0x0C10
IPSBAR + 0x0C14
IPSBAR + 0x0C18
IPSBAR + 0x0C1C–
IPSBAR + 0x0C3C
Table 15-3. Interrupt Controller Memory Map
Bits[31:24]
IRLR[7:1]
Bits[23:16] Bits[15:8]
Interrupt Pending Register High (IPRH), [63:32]
Interrupt Pending Register Low (IPRL), [31:1]
Interrupt Mask Register High (IMRH), [63:32]
Interrupt Mask Register Low (IMRL), [31:0]
Interrupt Force Register High (INTFRCH), [63:32]
Interrupt Force Register Low (INTFRCL), [31:1]
IACKLPR[7:0]
Reserved
Reserved
Bits[7:0]
15-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Interrupt Controller Module
Module Offset
IPSBAR + 0x0C40
IPSBAR + 0x0C44
IPSBAR + 0x0C48
IPSBAR + 0x0Cx4C
IPSBAR + 0x0C50
IPSBAR + 0x0C54
IPSBAR + 0x0C58
IPSBAR + 0x0C5C
IPSBAR + 0x0C60
IPSBAR + 0x0C64
IPSBAR + 0x0C68
IPSBAR + 0x0C6C
IPSBAR + 0x0C70
IPSBAR + 0x0C74
IPSBAR + 0x0C78
IPSBAR + 0x0C7C
IPSBAR + 0x0C80–
IPSBAR + 0x0CDC
IPSBAR + 0x0CE0
IPSBAR + 0x0CE4
IPSBAR + 0x0CE8
IPSBAR + 0x0CEC
IPSBAR + 0x0CF0
IPSBAR + 0x0CF4
IPSBAR + 0x0CF8
IPSBAR + 0x0CFC
Table 15-3. Interrupt Controller Memory Map (continued)
Bits[31:24]
ICR32
ICR36
ICR40
ICR44
ICR48
ICR52
ICR56
ICR60
Reserved
ICR04
ICR08
ICR12
ICR16
ICR20
ICR24
ICR28
Bits[23:16]
ICR33
ICR37
ICR41
ICR45
ICR49
ICR53
ICR57
ICR61
ICR01
ICR05
ICR09
ICR13
ICR17
ICR21
ICR25
ICR29
Reserved
Bits[15:8]
ICR34
ICR38
ICR42
ICR46
ICR50
ICR54
ICR58
ICR62
ICR02
ICR06
ICR10
ICR14
ICR18
ICR22
ICR26
ICR30
SWIACK
L1IACK
L2IACK
L3IACK
L4IACK
L5IACK
L6IACK
L7IACK
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
15.3
Register Descriptions
The interrupt controller registers are described in the following sections.
Bits[7:0]
ICR35
ICR39
ICR43
ICR47
ICR51
ICR55
ICR59
ICR63
ICR03
ICR07
ICR11
ICR15
ICR19
ICR23
ICR27
ICR31
15.3.1
Interrupt Pending Registers (IPRH
n
, IPRL
n
)
The IPRH n and IPRL n registers,
Figure 15-1 and Figure 15-2 , each 32 bits, provide a bit map for each
interrupt request to indicate if there is an active request (1 = active request, 0 = no request) for the given source. The state of the interrupt mask register does not affect the IPR n . The IPR n is cleared by reset. The
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
15-5
Interrupt Controller Module
IPR n is a read-only register, so any attempted write to this register is ignored. Bit 0 is not implemented and reads as a zero.
Access: read-only IPSBAR
Offset:
0xC00( (IPRH n )
30 31
R
W
Reset 0 0
29
0
28
0
27
0
26
0
25
0
24 23
INT[63:48]
0 0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
R
W
Reset 0
14
0
13 12 11 10 9 8 7
INT[47:32]
6 5 4 3
0
2
0
1
0
0
0 0 0 0 0 0 0 0 0 0 0
Figure 15-1. Interrupt Pending Register High (IPRH n )
Table 15-4. IPRH n Field Descriptions
Field Description
31–0
INT
Interrupt pending. Each bit corresponds to an interrupt source. The corresponding IMRH n bit determines whether an interrupt condition can generate an interrupt. At every system clock, the IPRH n samples the signal generated by the interrupting source. The corresponding IPRH n bit reflects the state of the interrupt signal even if the corresponding
IMRH n bit is set.
0 The corresponding interrupt source does not have an interrupt pending
1 The corresponding interrupt source has an interrupt pending
IPSBAR
Offset:
0xC04 (IPSBMT)
30 29 31
R
W
Reset 0 0 0
15
R
W
Reset 0
14
0
13
0
28
0
12
27
0
11
26
0
10
25 24 23
INT[31:16]
22
0 0 0
9 8
INT[16:1]
7
0
6
21
0
5
20
0 0 0 0 0 0 0 0
Figure 15-2. Interrupt Pending Register Low (IPRL n )
0
0
4
19 18
0
3
0
0
2
0
Access: read-only
17
0
1
0
16
0
0
0
0
15-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Interrupt Controller Module
Table 15-5. IPRL n Field Descriptions
Field Description
31–1
INT
Interrupt Pending. Each bit corresponds to an interrupt source. The corresponding IMRL n bit determines whether an interrupt condition can generate an interrupt. At every system clock, the IPRL n samples the signal generated by the interrupting source. The corresponding IPRL n bit reflects the state of the interrupt signal even if the corresponding
IMRL n bit is set.
0 The corresponding interrupt source does not have an interrupt pending
1 The corresponding interrupt source has an interrupt pending
0 Reserved, should be cleared.
15.3.2
Interrupt Mask Register (IMRH
n
, IMRL
n
)
The IMRH n and IMRL n registers are each 32 bits and provide a bit map for each interrupt to allow the request to be disabled (1 = disable the request, 0 = enable the request). The IMR n is set to all ones by reset, disabling all interrupt requests. The IMR n can be read and written. A write that sets bit 0 of the IMR forces the other 63 bits to be set, disabling all interrupt sources, and providing a global mask-all capability.
Access: read/write IPSBAR
Offset:
0xC08( (IMRH n )
30 31
R
W
Reset 1 1
29
1
28
1
27
1
26
1
25 24 23
INT_MASK[63:48]
22
1 1 1 1
21
1
20
1
19
1
18
1
17
1
16
1
15
R
W
Reset 1
14
1
13
1
12 11 10 9 8 7
INT_MASK[47:32]
6 5 4
1 1 1 1 1 1 1 1
Figure 15-3. Interrupt Mask Register High (IMRH n )
1
3
1
2
1
1
1
0
1
Table 15-6. IMRH n Field Descriptions
Field Description
31–0
INT_MASK
Interrupt mask. Each bit corresponds to an interrupt source. The corresponding IMRH n bit determines whether an interrupt condition can generate an interrupt. The corresponding IPRH n bit reflects the state of the interrupt signal even if the corresponding IMRH n bit is set.
0 The corresponding interrupt source is not masked
1 The corresponding interrupt source is masked
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
15-7
Interrupt Controller Module
IPSBAR
Offset:
0xC0C( (IMRL n )
30 31
R
W
Reset 1 1
29
1
15
R
W
Reset 1
14
1
13
1
28 27 26 25 24 23
INT_MASK[31:16]
22
1 1 1 1
21 20
1 1 1 1 1
12 11 10 9 8 7
INT_MASK[16:1]
6 5 4
1 1 1 1 1 1 1 1
Figure 15-4. Interrupt Mask Register Low (IMRL n )
1
19 18
1
3
1 1
1
2
Access: read/write
17 16
1
1
1
1
0
MAS
KALL
1
Table 15-7. IMRL n Field Descriptions
Field Description
31–1
INT_MASK
Interrupt mask. Each bit corresponds to an interrupt source. The corresponding IMRL n bit determines whether an interrupt condition can generate an interrupt. The corresponding IPRL n bit reflects the state of the interrupt signal even if the corresponding IMRL n bit is set.
0 The corresponding interrupt source is not masked
1 The corresponding interrupt source is masked
0
MASKALL
Mask all interrupts. Setting this bit will force the other 63 bits of the IMRH n and IMRL n to ones, disabling all interrupt sources, and providing a global mask-all capability.
NOTE
A spurious interrupt may occur if an interrupt source is being masked in the interrupt controller mask register (IMR) or a module’s interrupt mask register while the interrupt mask in the status register (SR[I]) is set to a value lower than the interrupt’s level. This is because by the time the status register acknowledges this interrupt, the interrupt has been masked. A spurious interrupt is generated because the CPU cannot determine the interrupt source.
To avoid this situation for interrupts sources with levels 1–6, first write a higher level interrupt mask to the status register, before setting the mask in the IMR or the module’s interrupt mask register. After the mask is set, return the interrupt mask in the status register to its previous value. Since level 7 interrupts cannot be disabled in the status register prior to masking, use of the IMR or module interrupt mask registers to disable level seven interrupts is not recommended.
15.3.3
Interrupt Force Registers (INTFRCH
n
, INTFRCL
n
)
The INTFRCH n and INTFRCL n registers, each 32 bits, provide a mechanism to allow software generation of interrupts for each possible source for functional or debug purposes. The system design may reserve one
15-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Interrupt Controller Module or more sources to allow software to self-schedule interrupts by forcing one or more of these bits (1 = force request, 0 = negate request) in the appropriate INTFRC n register. The assertion of an interrupt request via the INTFRC n register is not affected by the interrupt mask register. The INTFRC n register is cleared by reset.
Access: read/write IPSBAR
Offset:
0xC10( (INTFRCH n )
30 29 31
R
W
Reset 0 0 0
28
0
27
0
26
0
25 24 23
INTFRCH[63:48]
22
0 0 0 0
21
0
20
0
19
0
18
0
17
0
16
0
15
R
W
Reset 0
14
0
13 12 11 10 9 8 7
INTFRCH[47:32]
6 5 4
0 0 0 0 0 0 0 0 0 0
Figure 15-5. Interrupt Force Register High (INTFRCH n )
3
0
2
0
1
0
0
0
Table 15-8. INTFRCH n Field Descriptions
Field Description
31–0
INTFRCH
Interrupt force. Allows software generation of interrupts for each possible source for functional or debug purposes.
0 No interrupt forced on corresponding interrupt source
1 Force an interrupt on the corresponding source
IPSBAR
Offset:
0xC14( (INTFRCL n )
30 29 31
R
W
Reset 0 0 0
28
0
15
R
W
Reset 0
14
0
13 12
27
0
11
26
0
10
25 24 23
INTFRCL[31:16]
22
0 0 0 0
9 8 7
INTFRCL[16:1]
6
21
0
5
20
0
4
0 0 0 0 0 0 0 0 0 0
Figure 15-6. Interrupt Force Register Low (INTFRCL n )
19 18
0
3
0
0
2
0
Access: read/write
17
0
1
0
16
0
0
0
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
15-9
Interrupt Controller Module
.
31 16
Field
Reset
R/W
INTFRCL[31:16]
0000_0000_0000_0000
R/W
Field
Reset
R/W
15
INTFRCL[16:1]
0000_0000_0000_0000
R/W
IPSBAR + 0xC14
Figure 15-7. Interrupt Force Register Low (INTFRCL n )
Table 15-9. INTFRCL n Field Descriptions
1 0
—
Field Description
31–1
INTFRCL
Interrupt force. Allows software generation of interrupts for each possible source for functional or debug purposes.
0 No interrupt forced on corresponding interrupt source
1 Force an interrupt on the corresponding source
0 Reserved, should be cleared.
read/
15.3.4
Interrupt Request Level Register (IRLR
n
)
This 7-bit register is updated each machine cycle and represents the current interrupt requests for each interrupt level, where bit 7 corresponds to level 7, bit 6 to level 6, etc.
IPSBAR
Offset:
0xC18 (IRLR n )
7
R
W
Reset: 0
6 5 4
IRQ[7:1]
3 2
0 0 0 0 0
Figure 15-8. Interrupt Request Level Register (IRLR n )
1
0
Access: read-only
0
0
0
15-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Interrupt Controller Module
Table 15-10. IRLR n Field Descriptions
Field Description
7–1
IRQ
0
Interrupt requests. Represents the prioritized active interrupts for each level.
0 There are no active interrupts at this level
1 There is an active interrupt at this level
Reserved
15.3.5
Interrupt Acknowledge Level and Priority Register (IACKLPR
n
)
Each time an IACK is performed, the interrupt controller responds with the vector number of the highest priority source within the level being acknowledged. In addition to providing the vector number directly for the byte-sized IACK read, this 8-bit register is also loaded with information about the interrupt level and priority being acknowledged. This register provides the association between the acknowledged physical interrupt request number and the programmed interrupt level/priority. The contents of this
read-only register are described in Figure 15-9 and Table 15-11
.
Access: read-only IPSBAR
Offset:
0xC19 (IACKLPR n )
7
0 R
W
Reset: 0
6 5
LEVEL
4 3 2
PRI
0 0 0 0 0
Figure 15-9. IACK Level and Priority Register (IACKLPR n )
1
0
0
0
Table 15-11. IACKLPR n Field Descriptions
Field Description
7 Reserved
6–4
LEVEL
Interrupt level. Represents the interrupt level currently being acknowledged.
3–0
PRI
Interrupt Priority. Represents the priority within the interrupt level of the interrupt currently being acknowledged.
0 Priority 0
1 Priority 1
2 Priority 2
3 Priority 3
4 Priority 4
5 Priority 5
6 Priority 6
7 Priority 7
8 Mid-Point Priority associated with the fixed level interrupts only
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
15-11
Interrupt Controller Module
15.3.6
Interrupt Control Register (ICR
nx
)
Each ICR nx , where x = 1, 2,..., 63, specifies the interrupt level (1–7) and the priority within the level (0–7).
All ICR nx registers can be read, but only ICR n 8 to ICR n 63 can be written. It is software’s responsibility to program the ICR nx registers with unique and non-overlapping level and priority definitions. Failure to program the ICR nx registers in this manner can result in undefined behavior. If a specific interrupt request is completely unused, the ICR nx value can remain in its reset (and disabled) state.
IPSBAR
Offset:
and
for register offsets (ICR nx )
7
0
6
0
5 4
R
W
Reset: 0
IL
3
Access: R/W (Read only for ICR n 1-ICR n 7)
2
0 0 0 0
Figure 15-10. Interrupt Control Register (ICR nx )
0
1
IP
0
0
0
Table 15-12. ICR nx Field Descriptions
Field Description
7–6 Reserved, should be cleared.
5–3
IL
Interrupt level. Indicates the interrupt level assigned to each interrupt input.
2–0
IP
Interrupt priority. Indicates the interrupt priority for internal modules within the interrupt-level assignment. 000b represents the lowest priority and 111b represents the highest. For the fixed level interrupt sources, the priority is fixed at the midpoint for the level, and the IP field will always read as 000b.
15-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Interrupt Controller Module
18
19
20
21
22
Source Module
13
14
15
16
9
10
11
12
7
8
5
6
3
4
1
2
EPORT
SCM
DMA
UART0
UART1
UART2
17 I
2
C
QSPI
DTIM0
DTIM1
DTIM2
DTIM3
15.3.6.1
Interrupt Sources
shows the interrupt sources for each interrupt request line on Interrupt Controller 0.
Table 15-13. Interrupt Source Assignment For Interrupt Controller 0
Flag Source Description Flag Clearing Mechanism
IIF
INT
INT
INT
INT
INT
EPF1
EPF2
EPF3
EPF4
EPF5
EPF6
EPF7
SWTI
Edge port flag 1
Edge port flag 2
Edge port flag 3
Edge port flag 4
Edge port flag 5
Edge port flag 6
Edge port flag 7
Software watchdog timeout
Write EPF1 = 1
Write EPF2 = 1
Write EPF3 = 1
Write EPF4 = 1
Write EPF5 = 1
Write EPF6 = 1
Write EPF7 = 1
Cleared when service complete.
DONE DMA Channel 0 transfer complete Write DONE = 1
DONE DMA Channel 1 transfer complete Write DONE = 1
DONE DMA Channel 2 transfer complete Write DONE = 1
DONE DMA Channel 3 transfer complete Write DONE = 1
INT
INT
INT
UART0 interrupt
UART1 interrupt
UART2 interrupt
Automatically cleared
Automatically cleared
Automatically cleared
Not used (Reserved)
I
2
C interrupt
QSPI interrupt
DTIM0 interrupt
DTIM1 interrupt
DTIM2 interrupt
DTIM3 interrupt
Write IIF = 0
Write 1 to appropriate QIR bit
Write 1 to appropriate DTER0 bit
Write 1 to appropriate DTER1 bit
Write 1 to appropriate DTER2 bit
Write 1 to appropriate DTER3 bit
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
15-13
Interrupt Controller Module
Source Module
51
52
53
47
48
49
50
43
44
45
46
39
40
41
42
35
36
37
38
31
32
33
34
27
28
29
30
23
24
25
26
FEC
EPHY
GPT
PMM
ADC
PWM
RNGA
Table 15-13. Interrupt Source Assignment For Interrupt Controller 0 (continued)
Flag Source Description Flag Clearing Mechanism
X_INTF Transmit frame interrupt
X_INTB Transmit buffer interrupt
UN
RL
Transmit FIFO underrun
Collision retry limit
R_INTF Receive frame interrupt
RINTB Receive buffer interrupt
MII
LC
MII interrupt
Late Collision
HBERR Heartbeat error
GRA Graceful stop complete
EBERR Ethernet bus error
BABT Babbling transmit error
BABR Babbling receive error
EPHY EPHY interrupt
Write X_INTF = 1
Write X_INTB = 1
Write UN = 1
Write RL = 1
Write R_INTF = 1
Write R_INTB = 1
Write MII = 1
Write LC = 1
Write HBERR = 1
Write GRA = 1
Write EBERR = 1
Write BABT = 1
Not Used
Not Used
Write BABR = 1
Write EPHYIF = 1
TOF
PAIF
Timer overflow
Pulse accumulator input
PAOVF Pulse accumulator overflow
C0F Timer channel 0
Not Used
Not Used
Write TOF = 1 or access TIMCNTH/L if TFFCA = 1
Write PAIF = 1 or access PAC if TFFCA = 1
Write PAOVF = 1 or access PAC if TFFCA = 1
Write C0F = 1 or access IC/OC if TFFCA = 1
C1F
C2F
Timer channel 1
Timer channel 2
C3F Timer channel 3
LVDF LVD
Write 1 to C1F or access IC/OC if TFFCA = 1
Write 1 to C2F or access IC/OC if TFFCA = 1
Write 1 to C3F or access IC/OC if TFFCA = 1
Write LVDF = 1
ADCA ADCA conversion complete
ADCB ADCB conversion complete
ADCINT ADC Interrupt
PWM PWM Interrupt
RNGA RNGA Interrupt
Write 1 to EOSI0
Write 1 to EOSI1
Write 1 to ZCI, LLMTI and HLMTI
Write PWMIF = 1
Clears after one cycle of module reset
15-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Interrupt Controller Module
Table 15-13. Interrupt Source Assignment For Interrupt Controller 0 (continued)
Source Module
58
59
60
61
54
55
56
57
62
63
PIT0
PIT1
CFM
CFM
CFM
CFM
RTC
Flag Source Description Flag Clearing Mechanism
PIF
PIF
PIT interrupt flag
PIT interrupt flag
Not used (Reserved)
Write PIF = 1 or write PMR
Write PIF = 1 or write PMR
Not Used (Reserved)
Not Used (Reserved)
CBEIF SGFM buffer empty
CCIF
PVIF
SGFM command complete
Protection violation
Write CBEIF = 1
Cleared automatically
Cleared automatically
AEIF
RTC
Access error
RTC Interrupt
Cleared automatically
Write 1 to appropriate bit.
4
5
2
3
6
7
.
shows the interrupt sources for each interrupt request line on Interrupt Controller 1
Table 15-14. Interrupt Source Assignment For Interrupt Controller 1
Source Module
1
Flag Source Description
Not Used
Flag Clearing Mechanism
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
15-15
Interrupt Controller Module
36
37
38
39
32
33
34
35
28
29
30
31
24
25
26
27
20
21
22
23
16
17
18
19
12
13
14
15
10
11
8
9
Source Module Flag Source Description
BUF0I Message Buffer 2 Interrupt
BUF1I Message Buffer 3 Interrupt
BUF2I Message Buffer 4 Interrupt
BUF3I Message Buffer 5 Interrupt
BUF4I Message Buffer 6 Interrupt
BUF5I Message Buffer 7 Interrupt
BUF6I Message Buffer 8 Interrupt
BUF7I Message Buffer 9 Interrupt
FLEXCAN
BUF8I Message Buffer 10 Interrupt
BUF9I Message Buffer 11 Interrupt
BUF10I Message Buffer 12 Interrupt
BUF11I Message Buffer 13 Interrupt
BUF12I Message Buffer 14 Interrupt
BUF13I Message Buffer 15 Interrupt
BUF14I Error Interrupt
BUF15I Message Buffer 2 Interrupt
ERR_INT Message Buffer 3 Interrupt
BOFF_INT Bus-Off Interrupt
EPORT
EPF0
EPF1
EPF2
EPF3
EPF4
EPF5
EPF6
EPF7
Edge port flag 0
Edge port flag 1
Edge port flag 2
Edge port flag 3
Edge port flag 4
Edge port flag 5
Edge port flag 6
Edge port flag 7
Write 1 to BUF2I after reading as 1
Write 1 to BUF3I after reading as 1
Write 1 to BUF4I after reading as 1
Write 1 to BUF5I after reading as 1
Write 1 to BUF6I after reading as 1
Write 1 to BUF7I after reading as 1
Write 1 to BUF8I after reading as 1
Write 1 to BUF9I after reading as 1
Write 1 to BUF10I after reading as 1
Write 1 to BUF11I after reading as 1
Write 1 to BUF12I after reading as 1
Write 1 to BUF13I after reading as 1
Write 1 to BUF14I after reading as 1
Write 1 to BUF15I after reading as 1
Read reported error bits in ESR or write 0 to ERR_INT
Write 1 to BUF2I after reading as 1
Write 1 to BUF3I after reading as 1
Write 0 to BOFF_INT
Not Used
Not Used
Not Used
Not Used
Not Used
Flag Clearing Mechanism
Not Used
Write EPF0 = 1
Write EPF1 = 1
Write EPF2 = 1
Write EPF3 = 1
Write EPF4 = 1
Write EPF5 = 1
Write EPF6 = 1
Write EPF7 = 1
15-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Source Module
52
53
54
55
48
49
50
51
44
45
46
47
40
41
42
43
60
61
62
63
56
57
58
59
Flag
Interrupt Controller Module
Flag Clearing Mechanism Source Description
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
Not Used
15.3.7
Software and Level
n
IACK Registers (SWIACK, L
n
IACK)
The 8 IACK registers can be explicitly addressed via the , or implicitly addressed via a processor-generated interrupt acknowledge cycle during exception processing. In either case, the interrupt controller’s actions are very similar.
When a leveln IACK arrives in the interrupt controller, the controller examines all the currently-active level n interrupt requests, determines the highest priority within the level, and then responds with the unique vector number corresponding to that specific interrupt source. The vector number is supplied as the data for the byte-sized IACK read cycle. In addition to providing the vector number, the interrupt controller
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
15-17
Interrupt Controller Module also loads the level and priority number for the level into the IACKLPR register, where it may be retrieved later.
This interrupt controller design also supports the concept of a software IACK. A software IACK allows an interrupt service routine to determine if there are other pending interrupts so that the overhead associated with interrupt exception processing (including machine state save/restore functions) can be minimized. In general, the software IACK is performed near the end of an interrupt service routine, and if there are additional active interrupt sources, the current interrupt service routine (ISR) passes control to the appropriate service routine, but without taking another interrupt exception.
When the interrupt controller receives a software IACK read, it returns the vector number associated with the highest level, highest priority unmasked interrupt source for that interrupt controller. The IACKLPR register is also loaded as the software IACK is performed. If there are no active sources, the interrupt controller returns an all-zero vector as the operand. For this situation, the IACKLPR register is also cleared.
In addition to the software IACK registers within each interrupt controller, there are global software IACK registers. A read from the global SWIACK will return the vector number for the highest level and priority unmasked interrupt source from all interrupt controllers. A read from one of the L n IACK registers will return the vector for the highest priority unmasked interrupt within a level for all interrupt controllers.
IPSBAR
Offset:
for register offsets
(SWIACK, L n IACK)
7 6 5 4
VECTOR
3
R
W
Reset:
2 1
0 0 0 0 0 0 0
Figure 15-11. Software and Level n IACK Registers (SWIACK, L n IACK)
Access: read-only
0
0
Table 15-15. SWIACK and L n IACK Field Descriptions
Field Description
7–0
VECTOR
Vector number. A read from the SWIACK register returns the vector number associated with the highest level, highest priority unmasked interrupt source. A read from one of the L n IACK registers returns the highest priority unmasked interrupt source within the level.
15.4
Low-Power Wakeup Operation
The system control module (SCM) contains an 8-bit low-power interrupt control register (LPICR) used explicitly for controlling the low-power stop mode. This register must explicitly be programmed by software to enter low-power mode.
15-18
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Interrupt Controller Module
The interrupt controller provides a special combinatorial logic path to provide a special wake-up signal to exit from the low-power stop mode. This special mode of operation works as follows:
1. LPICR[6:4] is loaded with the mask level that will be specified while the core is in stop mode.
LPICR[7] must be set to enable this mode of operation.
NOTE
The wakeup mask level taken from LPICR[6:4] is adjusted by hardware to allow a level 7 IRQ to generate a wakeup. That is, the wakeup mask value used by the interrupt controller must be in the range of 0–6.
2. The processor executes a STOP instruction which places it in stop mode. Once the processor is stopped, each interrupt controller enables a special logic path that evaluates the incoming interrupt sources in a purely combinatorial path; that is, there are no clocked storage elements. If an active interrupt request is asserted and the resulting interrupt level is greater than the mask value contained in LPICR[6:4], then each interrupt controller asserts the wake-up output signal, which is routed to the SCM and PLL module to re-enable the device’s clock trees and resume processing.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
15-19
Interrupt Controller Module
15-20
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 16
Edge Port Module (EPORT)
16.1
Introduction
Although this device has two edge port modules, the description included herein treats each module as a single entity. Pay particular attention to the note below, as the two modules are not completely identical.
Specifically, edge port module 0 has seven interrupt inputs while module 1 contains eight. The edge port module (EPORT) has seven external interrupt pins, IRQ7–IRQ1.
Each pin can be configured individually as a level-sensitive interrupt pin, an edge-detecting interrupt pin (rising edge, falling edge, or both), or a general-purpose input/output (I/O) pin.
NOTE
Not all EPORT signals may be output from the device. See Chapter 2,
“Signal Descriptions,” to determine which signals are available.
Stop
Mode
EPPAR[2n, 2n + 1]
Edge Detect
Logic
EPFR n D0
D1
Q D0
D1
Q
To Interrupt
Controller
EPPDR n
Synchronizer
EPIER n
Rising Edge of System Clock
EPDR n
EPDDR n
IRQ n pin
Figure 16-1. EPORT Block Diagram
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
16-1
Edge Port Module (EPORT)
NOTE
The GPIO module must be configured to enable the peripheral function of the appropriate pins (refer to Chapter 13, “General Purpose I/O Module” ) prior to configuring the edge port module.
16.2
Low-Power Mode Operation
This section describes the operation of the EPORT module in low-power modes. For more information on low-power modes, see Chapter 8, “Power Management.”
Table 16-1 shows EPORT module operation in
low-power modes, and describes how this module may exit from each mode.
NOTE
The low-power control register (CR) in the system control module specifies the interrupt level at or above which is needed to bring the device out of a low-power mode.
Table 16-1. Edge Port Module Operation in Low-power Modes
Low-power Mode
Wait
Doze
Stop
EPORT Operation
Normal
Normal
Level-sensing only
Mode Exit
Any IRQ n interrupt at or above level in LPCR
Any IRQ n interrupt at or above level in LPCR
Any IRQ n interrupt set for level-sensing at or above level in LPCR. See note below.
In wait and doze modes, the EPORT module continues to operate as it does in run mode. It may be configured to exit the low-power modes by generating an interrupt request on either a selected edge or a low level on an external pin. In stop mode, there are no clocks available to perform the edge-detect function. Only the level-detect logic is active (if configured) to allow any low level on the external interrupt pin to generate an interrupt (if enabled) to exit stop mode.
NOTE
In stop mode, the input pin synchronizer is bypassed for the level-detect logic since no clocks are available.
16.3
Interrupt/GPIO Pin Descriptions
All EPORT pins default to general-purpose input pins at reset. The pin value is synchronized to the rising edge of CLKOUT when read from the EPORT pin data register (EPPDR). The values used in the edge/level detect logic are also synchronized to the rising edge of CLKOUT. These pins use
Schmitt-triggered input buffers that have built-in hysteresis designed to decrease the probability of generating false, edge-triggered interrupts for slow rising and falling input signals.
When a pin is configured as an output, it is driven to a state whose level is determined by the corresponding bit in the EPORT data register (EPDR). All bits in the EPDR are set at reset.
16-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Edge Port Module (EPORT)
16.4
Memory Map/Register Definition
the EPORT memory map.EPORT0 has an IPSBAR offset base address of 0x0013_0000 while EPORT1 has an IPSBAR offset of 0x0014_0000.
Table 16-2. Edge Port Module Memory Map
IPSBAR
Offset
Register
Width
(bits)
Access Reset Value Section/Page
Supervisor Access Only Registers
1
EPORT Pin Assignment Register (EPPAR) 16 R/W 0x0000
0x13_0000
0x14_0000
0x13_0002
0x14_0002
0x13_0003
0x14_0003
EPORT Data Direction Register (EPDDR)
EPORT Interrupt Enable Register (EPIER)
8
8
R/W
R/W
Supervisor/User Access Registers
0x13_0004
0x14_0004
0x13_0005
0x14_0005
EPORT Data Register (EPDR)
EPORT Pin Data Register (EPPDR)
8
8
R/W
R
0x13_0006
0x14_0006
EPORT Flag Register (EPFR) 8 R/W
1
User access to supervisor only address locations have no effect and result in a bus error
0x00
0x00
0xFF
See Section
0x00
16.4.1
EPORT Pin Assignment Register (EPPAR)
The EPORT pin assignment register (EPPAR) controls the function of each pin individually.
Offset
Address:
0x13_0000 (EPPAR0)
0x14_0000 (EPPAR1)
13 15 14
R
W
Reset 0
EPPA7
0
12
EPPA6
11 10
EPPA5
9
EPPA4
8 7
EPPA3
6 5
EPPA2
4
0 0 0 0 0 0 0 0 0 0
Figure 16-2. EPORT Pin Assignment Register (EPPAR)
Access: Supervisor read/write
3
0
EPPA1
2
0
1
0
0
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
16-3
Edge Port Module (EPORT)
Table 16-3. EPPAR Field Descriptions
Field Description
15–2
EPPA n
1–0
EPORT pin assignment select fields. The read/write EPPA n fields configure EPORT pins for level detection and rising and/or falling edge detection.
Pins configured as level-sensitive are active-low (logic 0 on the external pin represents a valid interrupt request).
Level-sensitive interrupt inputs are not latched. To guarantee that a level-sensitive interrupt request is acknowledged, the interrupt source must keep the signal asserted until acknowledged by software. Level sensitivity must be selected to bring the device out of stop mode with an IRQ n interrupt.
Pins configured as edge-triggered are latched and need not remain asserted for interrupt generation. A pin configured for edge detection can trigger an interrupt regardless of its configuration as input or output.
Interrupt requests generated in the EPORT module can be masked by the interrupt controller module. EPPAR functionality is independent of the selected pin direction.
Reset clears the EPPA n fields.
00 Pin IRQ n level-sensitive
01 Pin IRQ n rising edge triggered
10 Pin IRQ n falling edge triggered
11 Pin IRQ n both falling edge and rising edge triggered
Reserved, should be cleared.
16.4.2
EPORT Data Direction Register (EPDDR)
The EPORT data direction register (EPDDR) controls the direction of each one of the pins individually.
Offset
Address:
0x13_0002 (EPDDR0)
0x14_0002 (EPDDR1)
13 15 14
R
W
EPDD7
Reset 0 0
EPDD6
0
12
0
11
EPDD5
0
10
0
9
EPDD4
0
8
0
7
EPDD3
0
6
0
5
EPDD2
0
4
0
Figure 16-3. EPORT Data Direction Register (EPDDR)
Access: Supervisor read/write
3
EPDD1
0
2
0
1
0
0
0
Table 16-4. EPDDR Field Descriptions
Field Description
7–1
EPDD n
Setting any bit in the EPDDR configures the corresponding pin as an output. Clearing any bit in EPDDR configures the corresponding pin as an input. Pin direction is independent of the level/edge detection configuration. Reset clears
EPDD7–EPDD1.
To use an EPORT pin as an external interrupt request source, its corresponding bit in EPDDR must be clear.
Software can generate interrupt requests by programming the EPORT data register when the EPDDR selects output.
0 Corresponding EPORT pin configured as input
1 Corresponding EPORT pin configured as output
0 Reserved, should be cleared.
16.4.3
Edge Port Interrupt Enable Register (EPIER)
The EPORT interrupt enable register (EPIER) enables interrupt requests for each pin individually.
16-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Edge Port Module (EPORT)
Offset
Address:
0x13_0003 (EPIER0)
0x14_0003 (EPIER1)
7 6
R
W
Reset:
EPIE7
0
EPIE6
5
EPIE5
4
EPIE4
3
EPIE3
2
EPIE2
0 0 0 0 0
Figure 16-4. EPORT Port Interrupt Enable Register (EPIER)
Access: User read/write
1
EPIE1
0
0
0
0
Table 16-5. EPIER Field Descriptions
Field Description
7–1
EPIE n
0
Edge port interrupt enable bits enable EPORT interrupt requests. If a bit in EPIER is set, EPORT generates an interrupt request when:
• The corresponding bit in the EPORT flag register (EPFR) is set or later becomes set.
• The corresponding pin level is low and the pin is configured for level-sensitive operation.
Clearing a bit in EPIER negates any interrupt request from the corresponding EPORT pin. Reset clears
EPIE7–EPIE1.
0 Interrupt requests from corresponding EPORT pin disabled
1 Interrupt requests from corresponding EPORT pin enabled
Reserved, should be cleared.
16.4.4
Edge Port Data Register (EPDR)
The EPORT data register (EPDR) holds the data to be driven to the pins.
Offset
Address:
0x13_0004 (EPDR0)
0x14_0004 (EPDR1)
7 6
R
W
Reset:
EPD7
1
EPD6
5
EPD5
4
EPD4
3
EPD3
2
EPD2
1 1 1 1
Figure 16-5. EPORT Port Data Register (EPDR)
1
Access: User read/write
1
EPD1
1
0
0
1
Table 16-6. EPDR Field Descriptions
Field
7–1
EPD n
0
Description
Edge port data bits. Data written to EPDR is stored in an internal register; if any pin of the port is configured as an output, the bit stored for that pin is driven onto the pin. Reading EDPR returns the data stored in the register. Reset sets EPD7–EPD1.
Reserved, should be cleared.
16.4.5
Edge Port Pin Data Register (EPPDR)
The EPORT pin data register (EPPDR) reflects the current state of the pins.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
16-5
Edge Port Module (EPORT)
Offset
Address:
0x13_0005 (EPPDR0)
0x14_0005 (EPPDR1)
7
EPPD7
6
EPPD6 R
W
Reset: [IRQ7]
5
EPPD5
4
EPPD4
3
EPPD3
2
EPPD2
[IRQ6] [IRQ5] [IRQ4] [IRQ3] [IRQ2]
Figure 16-6. EPORT Port Pin Data Register (EPPDR)
Table 16-7. EPPDR Field Descriptions
Access: User read-only
1
EPPD1
[IRQ1]
0
0
0
Field Description
7–1
EPPD n
Edge port pin data bits. The read-only EPPDR reflects the current state of the EPORT pins IRQ7–IRQ1. Writing to
EPPDR has no effect, and the write cycle terminates normally. Reset does not affect EPPDR.
0 Reserved, should be cleared.
16.4.6
Edge Port Flag Register (EPFR)
The EPORT flag register (EPFR) individually latches EPORT edge events.
Offset
Address:
0x13_0006 (EPFR0)
0x14_0006 (EPFR1)
7 6
R
W
Reset:
EPF7
0
EPF6
5
EPF5
4
EPF4
3
EPF3
2
EPF2
0 0 0 0
Figure 16-7. EPORT Port Flag Register (EPFR)
0
Access: User read/write
1
EPF1
0
0
0
0
Table 16-8. EPFR Field Descriptions
Field
7–1
EPF n
0
Description
Edge port flag bits. When an EPORT pin is configured for edge triggering, its corresponding read/write bit in EPFR indicates that the selected edge has been detected. Reset clears EPF7–EPF1.
Bits in this register are set when the selected edge is detected on the corresponding pin. A bit remains set until cleared by writing a 1 to it. Writing 0 has no effect. If a pin is configured as level-sensitive (EPPAR n = 00), pin transitions do not affect this register.
0 Selected edge for IRQ n pin has not been detected.
1 Selected edge for IRQ n pin has been detected.
Reserved, should be cleared.
16-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 17
ColdFire Flash Module (CFM)
17.1
Introduction
17.1.1
Overview
The ColdFire Flash Module (CFM) is a non-volatile memory (NVM) module for integration with a CPU.
The CFM provides 256 Kbytes of 32-bit Flash memory serving as electrically erasable and programmable, non-volatile memory. The Flash memory is ideal for program and data storage for single-chip applications, allowing for field reprogramming without requiring external programming voltage sources.
The common Flash bus interface executes read operations to the Flash memory using one or two system bus cycles to access each Flash physical block, with access latency depending on the factory setting of the
CLKSEL bits in the CFMCLKSEL register. Flash physical blocks are interleaved between odd and even addresses to form a Flash logical block. Interleaving allows back-to-back read operations to the Flash memory at an effective access rate of one system bus cycle per word after the initial two-cycle access if the CLKSEL bits are not set for single cycle access.
It is not possible to read from any Flash logical block while the same logical block is being erased, programmed, or verified. Flash logical blocks are divided into multiple logical pages that can be erased separately. An erased bit reads 1 and a programmed bit reads 0.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
17-1
ColdFire Flash Module (CFM)
COMMON FLASH BUS
EVEN ODD
COMMON FLASH BUS INTERFACE
EVEN BLOCK ODD BLOCK
ARRAY 0 ARRAY 1 ARRAY 2 ARRAY 3
FLASH MEMORY CONTROLLER
FLASH COMMAND CONTROLLER
INTERNAL FLASH BUS INTERFACE
INTERNAL FLASH BUS
Figure 17-1. CFM Block Diagram
17.1.2
Features
• 256 Kbytes of 32-bit Flash memory
• Automated program, erase, and verify operations
• Single power supply for program and erase operations
• Software programmable interrupts on command completion, access violations, or protection violations
• Fast page erase operation
• Fast word program operation
17-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Flash Module (CFM)
• Protection scheme to prevent accidental program or erase of Flash memory
• Access restriction control for supervisor/user and data/instruction operations
• Security feature to prevent unauthorized access to the Flash memory
17.2
External Signal Description
The CFM contains no signals that connect off-chip for the end customer.
17.3
Memory Map and Register Definition
This section describes the CFM memory map and registers.
17.3.1
Memory Map
The memory map for the CFM Flash memory is shown in Figure 17-2 . The starting address of the Flash
memory is determined by the Flash array base address as defined by the system level configuration. The
Flash memory map shows how a pair of 32-bit Flash physical blocks (even and odd) interleave every 4 bytes to form a contiguous memory space as follows:
Flash Block 0 includes byte addresses (PROGRAM_ARRAY_BASE+$0000_0000) to
(PROGRAM_ARRAY_BASE+$0003_FFFF).
(PROGRAM_ARRAY_BASE + $0003_FFFF)
•
•
•
256kBytes
(PROGRAM_ARRAY_BASE + $0000_0000)
Configuration Field
(PROGRAM_ARRAY_BASE+$400) to
(PROGRAM_ARRAY_BASE+$417)
BLOCK ODD (4Bytes)
BLOCK EVEN (4Bytes)
BLOCK ODD (4Bytes)
BLOCK EVEN (4Bytes)
Figure 17-2. CFM Flash Memory Map
The CFM has hardware interlocks that protect data from accidental corruption using program or erase operations. A flexible scheme allows the protection of any combination of Flash logical sectors as
control supervisor/user and data/instruction access to these Flash logical sectors.
Security information that allows the MCU to prevent intrusive access to the Flash memory is stored in the
Flash configuration field. The Flash configuration field is composed of 24 bytes of reserved memory space within the Flash memory, which contains information that determines the CFM protection and access restriction scheme out of reset. A description of each byte found in the Flash configuration field is given
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
17-3
ColdFire Flash Module (CFM)
Address Offset
(from PROGRAM_ARRAY_BASE)
$0400 - $0407
$0408 - $040B
$040C - $040F
$0410 - $0413
$0414 - $0417
Table 17-1. CFM Configuration Field
Size
(bytes)
8
4
4
4
4
Description
Backdoor Comparison Key
Flash Protection Bytes
(see Section 17.3.2.4, “CFMPROT —
Flash SUPV Access Bytes
(see Section 17.3.2.5, “CFMSACC —
CFM Supervisor Access Register ”)
Flash DATA Access Bytes
(see Section 17.3.2.6, “CFMDACC —
”)
Flash Security Word
(see
”)
Factory
Default
$FFFF_FFFF_FFFF_FFFF
$FFFF_FFFF
$FFFF_FFFF
$FFFF_FFFF
$FFFF_FFFF
The CFM contains a set of control and status registers located at the register base address as defined by the
system level configuration. A summary of the CFM registers is given in Table 17-2
.
Table 17-2. CFM Register Address Map
1
Address Offset
(from REGISTER_BASE)
$0x0000
$0x0004
$0x0008
$0x000C
$0x0010
$0x0014
$0x0018
$0x001C
$0x0020
$0x0024
$0x0028
$0x002C
$0x0030
$0x0034
$0x0038
$0x003C
$0x0040
$0x0044
$0x0048
Register Bits
31 - 24
CFMMCR
CFMUSTAT
CFMCMD
RESERVED
23 - 16
1
15 - 8
CFMCLKD
RESERVED
1
CFMSEC
RESERVED
1
CFMPROT
CFMSACC
CFMDACC
RESERVED
1
RESERVED
1
RESERVED
1
RESERVED
1
RESERVED
1
RESERVED
1
RESERVED
1
RESERVED
1
RESERVED
1
RESERVED
1
RESERVED
1
CFMCLKSEL
7 - 0
RESERVED
1
Access to reserved address locations generate a cycle termination transfer error.
17-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Flash Module (CFM)
17.3.2
Register Descriptions
17.3.2.1
CFMMCR — CFM Module Configuration Register
The CFMMCR register is used to configure and control the operation of the internal bus interface.
IPSBAR
Offset:
0x0000 ( CFMMCR ) Access: User read/write
15
R 0
W
Reset 0
14
0
13
0
12
0
11 10 9 8 7
0 LOCK PVIE AEIE CBEI
E
6 5
CCIE KEYA
CC
4
0
3
0
2
0
1
0
0
0
0 0 0 0 0 0 0 0 0 0 0
Figure 17-3. CFM Module Configuration Register (CFMMCR)
0 0 0 0
CFMMCR register bits [10:5] are readable and writable with restrictions, while the remaining bits read 0 and are not writable.
Field
15-11
10
LOCK
9
PVIE
8
AEIE
7
CBEIE
6
CCIE
Table 17-3. CFMMCR Field Descriptions
Description
Reserved, will read as 0
Write lock control. The LOCK bit is always readable and is set once.
1 = CFMPROT, CMFSACC, and CFMDACC registers are write-locked.
0 = CFMPROT, CMFSACC, and CFMDACC registers are writable.
Protection violation interrupt enable
The PVIE bit is always readable and writable. The PVIE bit enables an interrupt in case the protection violation flag, PVIOL in the CFMUSTAT register, is set.
1 = An interrupt will be requested whenever the PVIOL flag is set.
0 = PVIOL interrupt disabled.
Access error interrupt enable
The AEIE bit is always readable and writable. The AEIE bit enables an interrupt in case the access error flag, ACCERR in the CFMUSTAT register, is set.
1 = An interrupt will be requested whenever the ACCERR flag is set.
0 = ACCERR interrupt disabled.
Command buffer empty interrupt enable
The CBEIE bit is always readable and writable. The CBEIE bit enables an interrupt in case the command buffer empty flag, CBEIF in the CFMUSTAT register, is set.
1 = An interrupt will be requested whenever the CBEIF flag is set.
0 = CBEIF interrupt disabled.
Command complete interrupt enable
The CCIE bit is always readable and writable. The CCIE bit enables an interrupt in case the command completion flag, CCIF in the CFMUSTAT register, is set.
1 = An interrupt will be requested whenever the CCIF flag is set.
0 = CCIF interrupt disabled.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
17-5
ColdFire Flash Module (CFM)
Field
5
KEYACC
4-0-
Table 17-3. CFMMCR Field Descriptions (continued)
Description
Enable security key writing
The KEYACC bit is readable and only writable if the KEYEN bits in the CFMSEC register are set to enable backdoor key access.
1 = Writes to CFM Flash memory are interpreted as keys to release security.
0 = Writes to CFM Flash memory are interpreted as the start of a command write sequence.
Reserved, will read as 0
17.3.2.2
CFMCLKD — CFM Clock Divider Register
The CFMCLKD register is used to control the period of the clock used for timed events in program and erase algorithms.
IPSBAR
Offset:
0x0002 (CFMCLKD) Access: User read/write
R
W
Reset:
7
DIVLD
6
PRDIV8
5 4 3
DIV
2 1
0 0 0 0 0 0 0
Figure 17-4. CFM Clock Divider Register (CFMCLKD)
All CFMCLKD register bits are readable, while bits [6:0] write once and bit 7 is not writable.
Table 17-4. CFMCLKD Field Descriptions
0
0
Field
7
DIVLD
6
PRDIV8
5-0
DIV
Description
Clock divider loaded
1 = CFMCLKD register has been written to since the last reset.
0 = CFMCLKD register has not been written.
Enable prescalar by 8
1 = Enables a prescalar to divide the internal Flash bus clock by 8 before feeding into the clock divider.
0 = The internal Flash bus clock is directly fed into the clock divider.
Clock divider bits
The combination of PRDIV8 and DIV effectively divides the internal Flash bus clock down to a frequency of 150 KHz - 200 KHz. The internal Flash bus clock frequency range is 150 KHz < internal Flash bus clock
< 102.4 MHz.
The CFMCLKD register bits PRDIV8 and DIV must be set with appropriate values before programming
or erasing the CFM Flash memory Section 17.4.2.3.1, “Writing the CFMCLKD Register
.”
17.3.2.3
CFMSEC — CFM Security Register
The CFMSEC register is used to store the Flash security word and CFM security state.
17-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Flash Module (CFM)
1
2
IPSBAR
Offset:
0x0008 (CFMSEC)
31 30 29
R KEYEN SECSTAT 0
W
Reset F
1
-
2
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
0 0 0 0 0 0 0 0
R
W
Reset
15 14 13 12 11 10 9 8
SEC
7 6 5
0 0 0 0 0 0 0 0 0 0 0
Figure 17-5. CFM Security Register (CFMSEC)
Reset state loaded from Flash configuration field during reset.
Reset state determined by security state of CFM.
20
0
0
4
0
Access: User read/write
19
0
0
3
0
18
0
0
2
0
17
0
0
1
0
16
0
0
0
0
CFMSEC register bits [31:30,15:0] are readable, while remaining bits read 0 and all bits are not writable.
Field
31
KEYEN
30
SECSTAT
29-16
15 - 0
SEC
Table 17-5.
Description
Enable backdoor key access to unlock security
1 = Backdoor key access to Flash module is enabled.
0 = Backdoor key access to Flash module is disabled.
Flash memory security status
1 = Flash security is enabled.
0 = Flash security is disabled.
Reserved, should read 0
Flash memory security bits
The SEC bits define the security state of the MCU as shown in
, which defines the single code that enables the security feature in the CFM
The CFMSEC register is loaded from the Flash configuration field in the Flash block at offset $0414 during the reset sequence, indicated by “F” in Figure 16-5 .
Table 17-6. CFM Security States
• SEC[15:0]
$4AC8
1
• Description
Flash Memory Secured
1
All other combinations Flash Memory Unsecured
This value was chosen because it represents the ColdFire HALT instruction, making it unlikely that a user compiled code accidentally programmed at the security configuration field location would unintentionally secure the Flash memory.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
17-7
ColdFire Flash Module (CFM)
The CFM Flash security operation is described in
Section 17.4.3, “Flash Security Operation
.
17.3.2.4
CFMPROT — CFM Protection Register
The CFMPROT register defines which Flash logical sectors are protected against program and erase operations.
IPSBAR
Offset:
0x0010 (CFMPROT) Access: User read/write
22 21 20 19 18 17 16
1
31 30 29 28 27 26 25 24 23
R
W
Reset F
1
F
1
F
1
F
1
F
1
F
1
F
1
PROTECT
F
1
F
1
15 14 13 12 11 10 9 8 7
R
W
Reset F
1
F
1
F
1
F
1
F
1
F
1
F
1
PROTECT
Reset state loaded from Flash configuration field during reset.
F
1
F
1
F
1
6
F
1
F
1
5
F
1
F
1
4
F
1
F
1
3
F
1
F
1
2
F
1
F
1
1
F
1
F
1
0
F
1
Figure 17-6. CFM Protection Register (CFMPROT)
All CFMPROT register bits are readable and only writable when LOCK=0.
The Flash memory is divided into logical sectors for the purpose of data protection using the CFMPROT register. The Flash memory consists of thirty-two 8-Kbyte sectors, as shown in
.
In order to change the Flash memory protection on a temporary basis, the CFMPROT register should be written after the LOCK bit in the CFMMCR register has been cleared. To change the Flash memory protection that will be loaded during the reset sequence, the Flash logical sector containing the Flash configuration field must first be unprotected, then the Flash protection bytes must be programmed with the desired value.
Table 17-7.
Field
31 - 0
PROTECT
Description
Each Flash logical sector can be protected from program and erase operations by setting the
PROTECT[M] bit.
PROTECT[M] = 1: Flash logical sector M is protected.
PROTECT[M] = 0: Flash logical sector M is not protected.
17-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Flash Module (CFM)
PROTECT[31:0] —
(PROGRAM_ARRAY_BASE + $0003_FFFF)
PROTECT[31]
SECTOR 31
(PROGRAM_ARRAY_BASE + $0003_E000)
(PROGRAM_ARRAY_BASE + $0002_2000)
•
•
•
SECTOR 16
(PROGRAM_ARRAY_BASE + $0002_0000)
SECTOR 15
(PROGRAM_ARRAY_BASE + $0001_E000)
(PROGRAM_ARRAY_BASE + $0000_2000)
PROTECT[0]
•
•
•
SECTOR 0
(PROGRAM_ARRAY_BASE + $0000_0000)
8kBytes
Flash Logical Sectors -
8kBytes
protection defined by CFMPROT
8kBytes
8kBytes
Figure 17-7. CFMPROT Protection Diagram
17.3.2.5
CFMSACC — CFM Supervisor Access Register
The CFMSACC register is used to control supervisor/user access to the Flash memory.
IPSBAR
Offset:
0x0014 (CFMSACC) Access: User read/write
23 22 21 20 19 18 17 16
1
31 30 29 28 27 26 25 24
R
W
Reset F
1
F
1
F
1
F
1
F
1
F
1
F
1
SUPV
F
1
15 14 13 12 11 10 9 8
R
W
Reset F
1
F
1
F
1
F
1
F
1
F
1
F
1
SUPV
Reset state loaded from Flash configuration field during reset.
F
1
F
1
7
F
1
F
1
6
F
1
F
1
5
F
1
F
1
4
F
1
F
1
3
F
1
F
1
2
F
1
F
1
1
F
1
F
1
0
F
1
Figure 17-8. CFM Supervisor Access Register (CFMSACC)
All CFMSACC register bits are readable and only writable when LOCK=0.
In order to change the Flash supervisor access on a temporary basis, the CFMSACC register should be written after the LOCK bit in the CFMMCR register has been cleared. To change the Flash supervisor access that will be loaded during the reset sequence, the Flash logical sector containing the Flash configuration field must first be unprotected, then the Flash supervisor access bytes must be programmed
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
17-9
ColdFire Flash Module (CFM) with the desired value. Each Flash logical sector may be mapped into supervisor or unrestricted address space (see
for details on Flash sector mapping).
Table 17-8.
Field
31 - 0
SUPV
Description
Flash address space assignment for supervisor/user access
SUPV[M] = 1: Flash logical sector M is placed in supervisor address space.
SUPV[M] = 0: Flash logical sector M is placed in unrestricted address space.
17.3.2.6
CFMDACC — CFM Data Access Register
The CFMDACC register is used to control data/instruction access to the Flash memory.
IPSBAR
Offset:
0x0018 (CFMDACC) Access: User read/write
23 22 21 20 19 18 17 16
1
31 30 29 28 27 26 25 24
R
W
Reset F
1
F
1
F
1
F
1
F
1
F
1
F
1
DACC
F
1
15 14 13 12 11 10 9 8
R
W
Reset F
1
F
1
F
1
F
1
F
1
F
1
F
1
DACC
Reset state loaded from Flash configuration field during reset.
F
1
F
1
7
F
1
F
1
6
F
1
F
1
5
F
1
F
1
4
F
1
F
1
3
F
1
F
1
2
F
1
F
1
1
F
1
F
1
0
F
1
Figure 17-9. CFM Data Access Register (CFMDACC)
All CFMDACC register bits are readable and only writable when LOCK=0.
In order to change the Flash data access on a temporary basis, the CFMDACC register should be written after the LOCK bit in the CFMMCR register has been cleared. To change the Flash data access that will be loaded during the reset sequence, the Flash logical sector containing the Flash configuration field must first be unprotected, then the Flash data access bytes must be programmed with the desired value. Each
for details on Flash sector mapping).
Table 17-9.
Field
31 - 0
DACC
Description
Flash memory address space assignment for data/instruction access
DACC[M] = 1: Flash logical sector M is placed in data address space.
DACC[M] = 0: Flash logical sector M is placed in data and instruction address space.
17-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Flash Module (CFM)
17.3.2.7
CFMUSTAT — CFM User Status Register
The CFMUSTAT register defines the Flash command controller status and Flash memory access, protection and verify status.
IPSBAR
Offset:
0x0020 (CFMUSTAT) Access: User read/write
7 6
CCIF
5 4 3
0
2 1
0
0
0 R
W
Reset:
CBEIF
1 1
PVIOL ACCERR
0 0 0
BLANK
0 0 0
Figure 17-10. CFM User Status Register (CFMUSTAT)
CFMUSTAT register bits CBEIF, PVIOL, ACCERR, and BLANK are readable and writable while CCIF is readable but not writable, and remaining bits read 0 and are not writable.
The CFMUSTAT register bits CBEIF, CCIF, PVIOL, ACCERR, and BLANK are available as external signals cfm_status_bits[7:4,2] on the module boundary.
NOTE
Only one CFMUSTAT register bit can be cleared at a time.
Table 17-10.
Field
7
CBEIF
6
CCIF
5
PVIOL
Description
Command buffer empty interrupt flag
The CBEIF flag, set by the Flash command controller, indicates that the address, data and command buffers are empty so that a new command write sequence can be started. The CBEIF flag is cleared by writing a 1 to CBEIF as part of a command write sequence. Writing a 0 to the CBEIF flag has no effect on CBEIF but can be used to abort a command write sequence. The CBEIF flag can generate an interrupt if the CBEIE bit in the CFMMCR register is set.
1 = Buffers are ready to accept a new command write sequence.
0 = Buffers are full.
Command complete interrupt flag
The CCIF flag, set by the Flash command controller, indicates that there are no more commands pending.
The CCIF flag is cleared by the Flash command controller when CBEIF is cleared and sets upon completion of all active and pending commands. Writing to the CCIF flag has no effect on CCIF. The CCIF flag can generate an interrupt if the CCIE bit in the CFMMCR register is set.
1 = All commands are completed.
0 = Command in progress.
Protection violation
The PVIOL flag, set by the Flash command controller, indicates an attempt was made to program or erase an address in a protected Flash logical sector. The PVIOL flag is cleared by writing a 1 to PVIOL. Writing a 0 to the PVIOL flag has no effect on PVIOL. While the PVIOL flag is set, it is not possible to launch a command or start a command write sequence.
1 = Protection violation has occurred.
0 = No protection violation has been detected.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
17-11
ColdFire Flash Module (CFM)
Field
4
ACCERR
3
2
BLANK
1 -0
Table 17-10.
Description
Access error
The ACCERR flag, set by the Flash command controller, indicates an illegal access was made to the
Flash memory or registers caused by an illegal command write sequence. The ACCERR flag is cleared by writing a 1 to the ACCERR flag. Writing a 0 to the ACCERR flag has no effect on ACCERR. While the
ACCERR flag is set, it is not possible to launch a command or start a command write sequence. See
Section 17.4.2.3.5, “Flash Normal Mode Illegal Operations
” for details on what action sets the ACCERR flag.
1 = Access error has occurred.
0 = No access error has been detected.
Reserved, should read 0
All Flash memory locations or the selected Flash logical page have been verified as erased
The BLANK flag, set by the Flash command controller, indicates that a blank check or page erase verify operation has checked all Flash memory locations or the selected Flash logical page and found them to be erased. The BLANK flag is cleared by writing a 1 to BLANK. Writing a 0 to the BLANK flag has no effect on BLANK.
1 = All Flash memory locations or selected logical page verify as erased.
0 = If a blank check or page erase verify command has been executed, and the CCIF flag is set, then a
0 in the BLANK flag indicates that all Flash memory locations are not erased or the selected Flash logical page is not erased.
Reserved, should read 0
17.3.2.8
CFMCMD — CFM Command Register
The CFMCMD register is the Flash command register.
IPSBAR
Offset:
0x0024 (CFMCMD)
7
0
6 5 4
R
W
Reset: 0 0 0 0
3
CMD
0
2
Access: User read/write
1 0
0 0 0
Figure 17-11. CFM Command Buffer and Register (CFMCMD)
All CFMCMD register bits are readable and writable except bit 7, which reads zero and is not writable.
Table 17-11.
Field
7
6 - 0
CMD
Description
Reserved, should read 0
Valid Flash memory commands are shown in
Table 17-12 . Writing a command other than those listed in
Table 17-12 during a command write sequence will cause the ACCERR flag in the CFMUSTAT register
to set.
17-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Flash Module (CFM)
Table 17-12. CFM Flash Memory Commands
CMD[6:0]
$05
$06
$20
$40
$41
Description
Blank Check
Page Erase Verify
Word Program
Page Erase
Mass Erase
17.3.2.9
CFMCLKSEL — CFM Clock Select Register
The CFMCLKSEL register reflects the factory setting for read access latency from the system bus to the
Flash block.
IPSBAR
Offset:
0x004A( CFMCLKSEL ) Access: User read/write
R
15
0
14
0
13
0
W
Reset 0 0 0
1
Reset state set by factory.
12
0
0
11
0
0
10
0
0
9
0
0
8
0
0
7
0
0
6
0
0
5
0
0
4
0
0
3
0
0
2
0
0
1 0
CLKSEL
F
1
F
1
Figure 17-12. CFM Clock Select Register (CFMCLKSEL)
CFMCLKSEL register bits [1:0] are read-only, while the remaining bits read 0 and are not writable.
Table 17-13.
Field
15 - 2
1 - 0
CLKSEL
Description
Reserved, should read 0
Flash read access latency select
The CLKSEL bits set the read access latency to the Flash block.
Table 17-14 describes the setting that
selects between single-cycle and two-cycle Flash block read access.
CLKSEL[1:0]
2’b10
All other combinations
Table 17-14. Clock Select States
Description
Single-Cycle Flash Block Read Access
Two-cycle Flash Block Read Access
Burst Read Access
1-1-1-1
2-1-1-1
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
17-13
ColdFire Flash Module (CFM)
17.4
Functional Description
17.4.1
General
The following modes and operations are described in the following sections:
1. Flash normal mode ( Section 17.4.2, “Flash Normal Mode ”)
a) Read operation (
Section 17.4.2.1, “Read Operation
”)
d) Stop mode (
Section 17.4.2.4, “Stop Mode ”)
2. Flash security operation ( Section 17.4.3, “Flash Security Operation ”)
17.4.2
Flash Normal Mode
In Flash normal mode, the user can access the CFM registers and the CFM Flash memory (see
Section 17.3.1, “Memory Map ”).
17.4.2.1
Read Operation
A valid read operation occurs whenever a transfer request is initiated, the address is equal to an address within the valid range of the CFM Flash memory space and the read/write control indicates a read cycle.
17.4.2.2
Write Operation
A valid write operation occurs whenever a transfer request is initiated, the address is equal to an address within the valid range of the CFM Flash memory space and the read/write control indicates a write cycle.
The action taken on a valid Flash array write depends on the subsequent user command issued as part of a valid command write sequence. Only 32-bit write operations are allowed to the Flash memory space. Byte and half-word write operations to the Flash memory space will result in a cycle termination transfer error.
17.4.2.3
Program, Erase, and Verify Operations
Write and read operations are both used for the program, erase, and verify algorithms described in this section. These algorithms are controlled by the Flash memory controller whose timebase, for program and erase operations, is derived from the internal Flash bus clock via a programmable counter.. The command register as well as the associated address and data registers operate as a buffer and a register (2-stage
FIFO), so that a new command along with the necessary data and address can be stored to the buffer while the previous command is still in progress. This buffering operation provides time optimization when programming more than one word on a physical row in the Flash memory as the high voltage generation can be kept active in between two programming operations, thereby saving the time overhead needed for setup of the high voltage charge pumps. Buffer empty as well as command completion are signaled by flags in the CFMUSTAT register with interrupts generated, if enabled.
The next four sections describe the following:
17-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Flash Module (CFM)
• How to write the CFMCLKD register
• Command write sequences used to program, erase, and verify the Flash memory
• Valid Flash commands
• Errors resulting from illegal command write sequences to the Flash memory
17.4.2.3.1
Writing the CFMCLKD Register
Prior to issuing any command, it is first necessary to write the CFMCLKD register to divide the input clock to within the 150 KHz to 200 KHz range. The CFMCLKD register bits PRDIV8 and DIV are set as follows:
For frequencies of the input clock greater than 12.8 MHz, the CFMCLKD bit PRDIV8 must be set.
CFMCLKD DIV bit field must be chosen such that the following equation is valid:
If PRDIV8 == 1 then FCLK = input clock / 8, else FCLK = input clock
If (FCLK[KHz] / 200KHz) is integer then DIV = (FCLK[KHz] / 200KHz) - 1, else DIV = INT (FCLK[KHz] / 200kHz)
FCLK, the clock to the Flash block timing control, is therefore:
FCLK = (input clock) / (DIV + 1)
150KHz < FCLK <= 200KHz
For example, if the input clock frequency is 33 MHz, the CFMCLKD DIV field should be set to $14 and bit PRDIV8 set to 1. The resulting FCLK is 196.4 KHz. As a result, the Flash memory program and erase algorithm timings are increased over the optimum target by:
(200 - 196.4) / 200 x 100% = 1.78%
Remark: INT(X) means taking the integer part of X
Example: INT(33MHz/8/200KHz) = 20
CAUTION
Programming the Flash with input clock < 150 KHz should be avoided.
Setting CFMCLKD to a value such that FCLK < 150 KHz can destroy the
Flash memory due to overstress. Setting CFMCLKD to a value such that
FCLK > 200 KHz can result in incomplete programming or erasure of the
Flash memory array cells.
NOTE
Program and Erase command execution time will increase proportionally with the period of FCLK.
If the CFMCLKD register is written, the DIVLD bit is set automatically. If the DIVLD bit is 0, the
CFMCLKD register has not been written since the last reset. No command can be executed if the
CFMCLKD register has not been written to Section 17.4.2.3.5, “Flash Normal Mode Illegal Operations
.”
17.4.2.3.2
Command Write Sequence
The Flash command controller is used to supervise the command write sequence to execute blank check, page erase verify, program, page erase, and mass erase algorithms.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
17-15
ColdFire Flash Module (CFM)
Before starting a command write sequence, the ACCERR and PVIOL flags in the CFMUSTAT register must be clear and the CBEIF flag should be tested to determine the state of the address, data, and command buffers. If the CBEIF flag is set, indicating the buffers are empty, a new command write sequence can be executed.
A command write sequence consists of three steps which must be strictly adhered to, because writes to the
CFM are not permitted between steps. However, Flash register and array reads are allowed during a command write sequence. The basic command write sequence is as follows:
1. Write to one or more addresses in the Flash memory.
2. Write a valid command to the CFMCMD register.
3. Clear CBEIF flag by writing a 1 to CBEIF to launch the command.
When the CBEIF flag is cleared, the CCIF flag will be cleared on the same bus cycle by the Flash command controller indicating that the command was successfully launched. The CBEIF flag will be set again indicating that the address, data, and command buffers are ready for a new command write sequence to begin. A buffered command will wait for the active command to be completed before being launched.
The CCIF flag in the CFMUSTAT register will set upon completion of all active and buffered commands.
A command write sequence can be aborted at anytime prior to clearing the CBEIF flag in the CFMUSTAT register by writing a 0 to the CBEIF flag. The ACCERR flag in the CFMUSTAT register will be set after successfully aborting a command write sequence and the ACCERR flag must be cleared prior to starting a new command write sequence.
17.4.2.3.3
Bus Arbitration During Write Operations
Once a command has been successfully launched, the CFM will signal the core platform to hold off read accesses to any active Flash physical block until all active and buffered commands have completed
(CCIF=1). A Flash write operation from the internal Flash bus will hold off the Core platform until it is completed.
17.4.2.3.4
Flash Normal Mode Commands
Table 17-15 summarizes the valid Flash normal mode commands.
Table 17-15. CFM Flash Memory Command Description
CFMCMD
$05
$06
$20
$40
$41
Meaning Description
Blank Check Verify that the entire Flash memory is erased. If all bits are erased, the BLANK bit will set
in the CFMUSTAT register, Figure 17-10 , upon command completion.
Page Erase
Verify
Program
Verifies that a Flash logical page is erased. If the Flash logical page is erased, the BLANK
bit will set in the CFMUSTAT register,
Program a 32-bit word.
Figure 17-10 , upon command completion.
Page Erase Erase a Flash logical page.
Mass Erase Erase the entire Flash memory. All Flash memory protection must be disabled.
17-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Flash Module (CFM)
Blank Check
The blank check operation will verify that all Flash memory addresses in the CFM are erased.
An example flow to execute the blank check command is shown in
. The blank check command write sequence is as follows:
1. Write to any Flash memory address to start the command write sequence for the blank check command. The specific address and data written during the blank check command write sequence will be ignored.
2. Write the blank check command, $05, to the CFMCMD register.
3. Clear the CBEIF flag by writing a 1 to CBEIF to launch the blank check command.
Because all Flash physical blocks are verified simultaneously, the number of internal Flash bus cycles required to execute the blank check operation on a fully erased Flash memory is equal to the number of word addresses in a Flash logical block plus 15 internal Flash bus cycles as measured from the time the
CBEIF flag is cleared until the CCIF flag is set in the CFMUSTAT register. Upon completion of the blank check operation (CCIF=1), the BLANK flag will set in the CFMUSTAT register if the entire Flash memory is erased. If any Flash memory location is not erased, the blank check operation will terminate and the
BLANK flag will remain clear.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
17-17
ColdFire Flash Module (CFM)
START
Clock Register
Written
Check
Read: Register CFMCLKD
Bit
DIVLD
Set?
yes no
Write: Register CFMCLKD
Read: Register CFMUSTAT
Address, Data,
Command
Buffer Empty Check
Bit
CBEIF
Set?
yes no
Access Error and
Protection Violation
Check
1.
Bit
ACCERR/PVIOL
Set?
yes Write: Register CFMUSTAT
Clear bit ACCERR/PVIOL $30 no
Write: Array Address
and Data
2.
3.
Write: Register CFMCMD
Blank Check Command $05
Write: Register CFMUSTAT
Clear bit CBEIF $80
NOTE: command write sequence aborted by writing $00 to
CFMUSTAT register.
NOTE: command write sequence aborted by writing $00 to
CFMUSTAT register.
Read: Register CFMUSTAT
Bit Polling for
Command Completion
Check
Bit
CCIF
Set?
no yes
Read: Register CFMUSTAT
Blank Check
Verify Status
Bit
BLANK
Set?
no yes
Write: Register CFMUSTAT
Clear bit BLANK $04
Flash Memory
Erased
EXIT EXIT
Flash Memory
Not Erased
Figure 17-13. Example Blank Check Command Flow
17-18
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Flash Module (CFM)
Page Erase Verify
The page erase verify operation will verify that all memory addresses in a Flash logical page are erased.
An example flow to execute the page erase verify operation is shown in Figure 17-14 . The page erase
verify command write sequence is as follows:
1. Write to any word address in a Flash logical page to start the command write sequence for the page erase verify command. The address written will determine the Flash logical page to be verified, while the data written during the page erase verify command write sequence will be ignored.
2. Write the page erase verify command, $06, to the CFMCMD register.
3. Clear the CBEIF flag by writing a 1 to CBEIF to launch the page erase verify command.
Since the word addresses in even and odd Flash blocks are interleaved, pages from adjacent interleaving
Flash physical blocks will automatically be erase verified at the same time. The number of internal Flash bus cycles required to execute the page erase verify operation on a fully erased Flash logical page is equal to the number of word addresses in a Flash logical page plus 15 internal Flash bus cycles as measured from the time the CBEIF flag is cleared until the CCIF flag is set in the CFMUSTAT register.
Upon completion of any page erase verify operation (CCIF=1), the BLANK flag in the CFMUSTAT register will be set if all addresses in the selected Flash logical page are verified to be erased. If any address in the selected Flash logical page is not erased, the page erase verify operation will terminate and the
BLANK flag will remain clear.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
17-19
ColdFire Flash Module (CFM)
START
Clock Register
Written
Check
Read: Register CFMCLKD
Bit
DIVLD
Set?
yes no
Write: Register CFMCLKD
Read: Register CFMUSTAT
Address, Data,
Command
Buffer Empty Check
Bit
CBEIF
Set?
yes no
Access Error and
Protection Violation
Check
1.
Bit
ACCERR/PVIOL
Set?
yes Write: Register CFMUSTAT
Clear bit ACCERR/PVIOL $30 no
Write: Logical Page Address
and Dummy Data
2.
3.
Write: Register CFMCMD
Page Erase Verify Command $06
Write: Register CFMUSTAT
Clear bit CBEIF $80
NOTE: command write sequence aborted by writing $00 to
CFMUSTAT register.
NOTE: command write sequence aborted by writing $00 to
CFMUSTAT register.
Read: Register CFMUSTAT
Bit Polling for
Command Completion
Check
Bit
CCIF
Set?
no yes
Read: Register CFMUSTAT
Page Erase
Verify Status
Bit
BLANK
Set?
no yes
Write: Register CFMUSTAT
Clear bit BLANK $04
Flash Logical Page
Erased
EXIT EXIT
Flash Logical Page
Not Erased
Figure 17-14. Example Page Erase Verify Command Flow
17-20
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Flash Module (CFM)
Program
The program operation will program a previously erased address in the Flash memory using an embedded algorithm.
An example flow to execute the program operation is shown in
. The program command write sequence is as follows:
1. Write to a word address in a Flash physical block to start the command write sequence for the program command. The word address written will determine the Flash physical block address to program while the data written during the program command write sequence will determine the data stored at that address. The same relative address in multiple program Flash physical blocks may be programmed simultaneously by writing to the relative address in Flash physical block order: even block, odd block. The Flash physical block written to in the first array write limits the ability to simultaneously program in block order only those Flash physical blocks that remain.
2. Write the program command, $20, to the CFMCMD register.
3. Clear the CBEIF flag by writing a 1 to CBEIF to launch the program command.
If the address to be programmed is in a protected sector of the Flash memory, the PVIOL flag in the
CFMUSTAT register will set and the program command will not launch. Once the program command has successfully launched, the CCIF flag in the CFMUSTAT register will set after the program operation has completed unless a new command write sequence has been buffered.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
17-21
ColdFire Flash Module (CFM)
START
Read: Register CFMCLKD
Clock Register
Written
Check
Bit
DIVLD
Set?
yes no
Write: Register CFMCLKD
Read: Register CFMUSTAT
Bit
CBEIF
Set?
yes no
Access Error and
Protection Violation
Check
1.
Bit
ACCERR/PVIOL
Set?
no yes Write: Register CFMUSTAT
Clear bit ACCERR/PVIOL $30
Write: Array Address and
Data
2.
3.
Write: Register CFMCMD
Program Command $20
Write: Register CFMUSTAT
Clear bit CBEIF $80
NOTE: command write sequence aborted by writing $00 to
CFMUSTAT register.
NOTE: command write sequence aborted by writing $00 to
CFMUSTAT register.
Read: Register CFMUSTAT
Protection
Violation Check
Address, Data,
Command
Buffer Empty Check
Bit
PVIOL
Set?
no
Bit
CBEIF
Set?
yes yes no
Read: Register CFMUSTAT
Write: Register CFMUSTAT
Clear bit PVIOL $20
Change Protection
•
Bit Polling for
Command
Completion Check
Bit
CCIF
Set?
yes
EXIT no yes
Next Write? no
Figure 17-15. Example Program Command Flow
Page Erase
The page erase operation will erase all memory addresses in a Flash logical page using an embedded algorithm.
17-22
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Flash Module (CFM)
An example flow to execute the page erase operation is shown in
. The page erase command write sequence is as follows:
1. Write to any word address in a Flash logical page to start the command write sequence for the page erase command. The word address written will determine the Flash logical page to erase while the data written during the page erase command write sequence will be ignored.
2. Write the page erase command, $40, to the CFMCMD register.
3. Clear the CBEIF flag by writing a 1 to CBEIF to launch the page erase command.
If the Flash logical page to be erased is in a protected sector of the Flash memory, the PVIOL flag in the
CFMUSTAT register will set and the page erase command will not launch. Once the page erase command has successfully launched, the CCIF flag in the CFMUSTAT register will set after the page erase operation has completed, unless a new command write sequence has been buffered.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
17-23
ColdFire Flash Module (CFM)
START
Read: Register CFMCLKD
Clock Register
Written
Check
Bit
DIVLD
Set?
yes no
Write: Register CFMCLKD
Read: Register CFMUSTAT
Bit
CBEIF
Set?
yes no
Access Error and
Protection Violation
Check
1.
Bit
ACCERR/PVIOL
Set?
no yes Write: Register CFMUSTAT
Clear bit ACCERR/PVIOL $30
Write: Logical Page Address
and Dummy Data
2.
3.
Write: Register CFMCMD
Page Erase Command $40
Write: Register CFMUSTAT
Clear bit CBEIF $80
NOTE: command write sequence aborted by writing $00 to
CFMUSTAT register.
NOTE: command write sequence aborted by writing $00 to
CFMUSTAT register.
Read: Register CFMUSTAT
Protection
Violation Check
Address, Data,
Command
Buffer Empty Check
Bit
PVIOL
Set?
no
Bit
CBEIF
Set?
yes yes no
Read: Register CFMUSTAT
Write: Register CFMUSTAT
Clear bit PVIOL $20
Change Protection
•
Bit Polling for
Command
Completion Check
Bit
CCIF
Set?
yes
EXIT no yes
Next Write? no
Figure 17-16. Example Page Erase Command Flow
Mass Erase
The mass erase operation will erase all Flash memory addresses using an embedded algorithm.
17-24
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Flash Module (CFM)
An example flow to execute the mass erase command is shown in
. The mass erase command write sequence is as follows:
1. Write to any Flash memory address to start the command write sequence for the mass erase command. The specific address and data written during the mass erase command write sequence will be ignored.
2. Write the mass erase command, $41, to the CFMCMD register.
3. Clear the CBEIF flag by writing a 1 to CBEIF to launch the mass erase command.
If any Flash logical sector is protected, the PVIOL flag in the CFMUSTAT register will set during the command write sequence and the mass erase command will not launch. Once the mass erase command has successfully launched, the CCIF flag in the CFMUSTAT register will set after the mass erase operation has completed, unless a new command write sequence has been buffered.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
17-25
ColdFire Flash Module (CFM)
START
Read: Register CFMCLKD
Clock Register
Written
Check
Bit
DIVLD
Set?
yes no
Write: Register CFMCLKD
Read: Register CFMUSTAT
Bit
CBEIF
Set?
yes no
Access Error and
Protection Violation
Check
1.
Bit
ACCERR/PVIOL
Set?
yes no
Write: Register CFMUSTAT
Clear bit ACCERR/PVIOL $30
Write: Array Address and
Dummy Data
2.
3.
Write: Register CFMCMD
Mass Erase Command $41
Write: Register CFMUSTAT
Clear bit CBEIF $80
NOTE: command write sequence aborted by writing $00 to
CFMUSTAT register.
NOTE: command write sequence aborted by writing $00 to
CFMUSTAT register.
Read: Register CFMUSTAT
Protection
Violation Check
Bit
PVIOL
Set?
no yes
Address, Data,
Command
Buffer Empty Check
Bit
CBEIF
Set?
yes no
Read: Register CFMUSTAT
Write: Register CFMUSTAT
Clear bit PVIOL $20
Change Protection
•
Bit Polling for
Command
Completion Check
Bit
CCIF
Set?
yes
EXIT no yes
Next Write? no
Figure 17-17. Example Mass Erase Command Flow
17-26
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ColdFire Flash Module (CFM)
17.4.2.3.5
Flash Normal Mode Illegal Operations
The ACCERR flag will be set during the command write sequence if any of the following illegal operations are performed, causing the command write sequence to immediately abort:
1. Writing to the Flash memory before initializing CFMCLKD.
2. Writing to the Flash memory while CBEIF is not set.
3. Writing to a Flash block with a data size other than 32 bits.
4. After writing to the Flash even block, writing an additional word to the Flash memory during the
Flash command write sequence other than the Flash odd block.
5. Writing an invalid Flash normal mode command to the CFMCMD register.
6. Writing to any CFM register other than CFMCMD after writing to the Flash memory.
7. Writing a second command to the CFMCMD register before executing the previously written command.
8. Writing to any CFM register other than CFMUSTAT (to clear CBEIF) after writing to the command register, CFMCMD.
9. The part enters stop mode and any command is in progress. Upon entering STOP mode, any active command is aborted.
10. Aborting a command write sequence by writing a 0 to the CBEIF flag after writing to the Flash memory or after writing a command to the CFMCMD register but before the command is launched.
The PVIOL flag will be set during the command write sequence if any of the following illegal operations are performed, causing the command write sequence to immediately abort:
1. Writing a program command if the address to program is in a protected Flash logical sector.
2. Writing a page erase command if the address to erase is in a protected Flash logical sector.
3. Writing a mass erase command while any protection is enabled .
If a read operation is attempted on a Flash logical block while a command is active on that logical block
(CCIF=0), the read operation will return invalid data and the ACCERR flag in the CFMUSTAT register will not be set.
17.4.2.4
Stop Mode
If a command is active (CCIF=0) when the MCU enters stop mode, the Flash command controller and
Flash memory controller will perform the following:
1. The active command will be aborted, and the data being programmed or erased is lost.
2. The high voltage circuitry to the Flash arrays will be switched off.
3. Any buffered command (CBEIF=0) will not be executed once the MCU exits stop mode.
4. The CCIF and ACCERR flags will be set if a command is active when the MCU enters stop mode.
CAUTION
As active commands are immediately aborted when the MCU enters stop mode, it is strongly recommended that the user does not execute the stop instruction during program and erase operations.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
17-27
ColdFire Flash Module (CFM)
If a command is not active (CCIF=1) when the MCU enters stop mode, the ACCERR flag will not set.
17.4.3
Flash Security Operation
The CFM provides security information to the Integration module and the rest of the MCU. This security information is stored within a word in the Flash configuration field. This security word is read automatically after each reset and stored in the CFMSEC register.
In Flash normal mode, the user can bypass the security via a backdoor access sequence using an 8-byte long key. Upon successful completion of the backdoor access sequence, the SECSTAT bit in the CFMSEC register is cleared indicating that the MCU is unsecured.
The CFM may be unsecured via one of the following methods:
1. Executing a backdoor access sequence.
2. Passing a blank check operation on the Flash memory.
17.4.3.1
Backdoor Access Sequence
If the KEYEN bits in the CFMSEC register are set to the enabled state, the user can bypass security by performing the following:
1. Setting the KEYACC bit in the CFMMCR register.
2. Writing the correct 8-byte backdoor comparison key to the Flash memory at offset $0400 - $0407.
This operation must be composed of two 32-bit writes to address $0400 and $0404 in that order.
The two backdoor write cycles can be separated by any number of internal Flash bus cycles.
NOTE
Any attempt to use a key of all zeros or all ones will lock the backdoor access sequence until the CFM is reset.
3. Clearing the KEYACC bit.
4. If all 8 bytes written match the Flash memory content at offset $0400 - $0407, then security is bypassed until the next reset.
In the unsecured state, the user has full control of the contents of the 8-byte backdoor comparison key by programming the bytes at offset $0400 - $0407 of the Flash configuration field. If at any time a key of all
0s or all 1s is received, the backdoor access sequence is terminated and cannot be successfully restarted until after the CFM is reset.
Note that the security of the CFM as defined in the Flash security word at address offset $0414 is not changed by the executing the backdoor access sequence to unsecure the device. After the next reset sequence, the CFM is secured again and the same backdoor key is in effect unless the Flash configuration field was changed by program or erase prior to reset. The backdoor access sequence to unsecure the device has no effect on the program and erase protections defined in the CFM protection register.
The contents of the Flash security word at address offset $0414 must be changed by programming that address when the device is unsecured and the sector containing the Flash configuration field is unprotected.
17-28
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
.
ColdFire Flash Module (CFM)
17.4.3.2
Blank Check
A secured CFM can be unsecured by verifying that the entire Flash memory is erased. If required, the mass erase command can be executed on the Flash memory. The blank check command must then be executed on the Flash memory. The CFM will be unsecured if the blank check operation determines that the entire
Flash memory is erased. After the next reset sequence, the security state of the CFM is determined by the
Flash security word at address offset $0414. For further details on security, see the MCU security specification.
17.4.3.3
JTAG Lockout Recovery
A secured CFM can be unsecured by mass erasing the Flash memory via a sequence of JTAG commands, as specified in the system level security documentation followed by a reset of the MCU.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
17-29
ColdFire Flash Module (CFM)
17-30
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 18
Fast Ethernet Controller (FEC)
This chapter provides a feature-set overview and a functional block diagram. Additionally, detailed descriptions of operation and the programming model are included.
18.1
Overview
The Ethernet Media Access Controller (MAC) is designed to support both 10 and 100 Mbps
Ethernet/IEEE 802.3 networks. The MAC must be used in conjunction with the on-board transceiver interface and transceiver function to complete the interface to the media.
18.1.1
Features
The FEC incorporates the following features:
• IEEE 802.3 full duplex flow control
• Programmable max frame length supports IEEE 802.1 VLAN tags and priority
• Support for full-duplex operation (200Mbps throughput) with a minimum system clock rate of
50MHz
• Support for half-duplex operation (100Mbps throughput) with a minimum system clock rate of 25
MHz
• Retransmission from transmit FIFO following a collision (no processor bus utilization)
• Automatic internal flushing of the receive FIFO for runts (collision fragments) and address recognition rejects (no processor bus utilization)
• Address recognition
— Frames with broadcast address may be always accepted or always rejected
— Exact match for single 48-bit individual (unicast) address
— Hash (64-bit hash) check of individual (unicast) addresses
— Hash (64-bit hash) check of group (multicast) addresses
— Promiscuous mode
18.2
Modes of Operation
The primary operational modes are described in this section.
18.2.1
Full and Half Duplex Operation
Full duplex mode is intended for use on point to point links between switches or end node to switch. Half duplex mode is used in connections between an end node and a repeater or between repeaters. Selection of the duplex mode is controlled by TCR[FDEN].
When configured for full duplex mode, flow control may be enabled. Refer to the TCR[RFC_PAUSE] and
TCR[TFC_PAUSE] bits, the RCR[FCE] bit, and
Section 18.4.10, “Full Duplex Flow Control ,” for more
details.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-1
Fast Ethernet Controller (FEC)
18.2.2
Interface Options
The following interface options are supported. A detailed discussion of the interface configurations is provided in
Section 18.4.5, “Network Interface Options
”.
18.2.2.1
10 Mbps and 100 Mbps MII Interface
MII is the Media Independent Interface defined by the IEEE 802.3 standard for 10/100 Mbps operation.
The MAC-PHY interface may be configured to operate in MII mode by asserting RCR[MII_MODE].
The speed of operation is determined by the ETXCLK and ERXCLK pins which are driven by the external transceiver. The transceiver will either auto-negotiate the speed or it may be controlled by software via the serial management interface (EMDC/EMDIO pins) to the transceiver. Refer to the MMFR and MSCR register descriptions as well as the section on the MII for a description of how to read and write registers in the transceiver via this interface.
18.2.3
Address Recognition Options
The address options supported are promiscuous, broadcast reject, individual address (hash or exact match),
”.
18.2.4
Internal Loopback
Internal loopback mode is selected via RCR[LOOP]. Loopback mode is discussed in detail in
Section 18.4.13, “Internal and External Loopback ”.
18-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
FEC Top-Level Functional Diagram
18.3
FEC Top-Level Functional Diagram
The block diagram of the FEC is shown below. The FEC is implemented with a combination of hardware and microcode. The off-chip (Ethernet) interfaces are compliant with industry and IEEE 802.3 standards.
SIF
Bus
Controller
Descriptor
Controller
(RISC + microcode)
CSR
RAM
RAM I/F
FIFO
Controller
DMA
MIB
Counters
Transmit Receive
FEC Bus
MII
MDO
MDEN
MDI
I/O
PAD
EMDIO EMDC
ETXEN
ETXD[3:0]
ETXER
ETCLK
ECRS,ECOL
MII/7-WIRE DATA
OPTION
Figure 18-1. FEC Block Diagram
ERXCLK
ERXDV
ERXD[3:0]
ERXER
The descriptor controller is a RISC-based controller that provides the following functions in the FEC:
• Initialization (those internal registers not initialized by the user or hardware)
• High level control of the DMA channels (initiating DMA transfers)
• Interpreting buffer descriptors
• Address recognition for receive frames
• Random number generation for transmit collision backoff timer
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-3
Fast Ethernet Controller (FEC)
NOTE
DMA references in this section refer to the FEC’s DMA engine. This DMA engine is for the transfer of FEC data only, and is not related to the DMA controller described in Chapter 16, “DMA Controller Module ,” nor to the
DMA timers described in Chapter 21, “DMA Timers (DTIM0–DTIM3) .”
The RAM is the focal point of all data flow in the Fast Ethernet Controller and is divided into transmit and receive FIFOs. The FIFO boundaries are programmable using the FRSR register. User data flows to/from the DMA block from/to the receive/transmit FIFOs. Transmit data flows from the transmit FIFO into the transmit block and receive data flows from the receive block into the receive FIFO.
The user controls the FEC by writing, through the SIF (Slave Interface) module, into control registers located in each block. The CSR (control and status register) block provides global control (e.g. Ethernet reset and enable) and interrupt handling registers.
The MII block provides a serial channel for control/status communication with the external physical layer device (transceiver). This serial channel consists of the EMDC (Management Data Clock) and EMDIO
(Management Data Input/Output) lines of the MII interface.
The DMA block provides multiple channels allowing transmit data, transmit descriptor, receive data and receive descriptor accesses to run independently.
The Transmit and Receive blocks provide the Ethernet MAC functionality (with some assist from microcode).
The Message Information Block (MIB) maintains counters for a variety of network events and statistics.
It is not necessary for operation of the FEC but provides valuable counters for network management. The counters supported are the RMON (RFC 1757) Ethernet Statistics group and some of the IEEE 802.3
counters. See Section 18.5.3, “MIB Block Counters Memory Map ” for more information.
18.4
Functional Description
This section describes the operation of the FEC, beginning with the hardware and software initialization sequence, then the software (Ethernet driver) interface for transmitting and receiving frames.
Following the software initialization and operation sections are sections providing a detailed description of the functions of the FEC.
18.4.1
Initialization Sequence
This section describes which registers are reset due to hardware reset, which are reset by the FEC RISC, and what locations the user must initialize prior to enabling the FEC.
18.4.1.1
Hardware Controlled Initialization
In the FEC, registers and control logic that generate interrupts are reset by hardware. A hardware reset deasserts output signals and resets general configuration bits.
Other registers reset when the ECR[ETHER_EN] bit is cleared. ECR[ETHER_EN] is deasserted by a hard reset or may be deasserted by software to halt operation. By deasserting ECR[ETHER_EN], the configuration control registers such as the TCR and RCR will not be reset, but the entire data path will be reset.
18-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Functional Description
Table 18-1. ECR[ETHER_EN] De-Assertion Effect on FEC
Register/Machine
XMIT block
RECV block
DMA block
RDAR
TDAR
Descriptor Controller block
Reset Value
Transmission is aborted (bad CRC appended)
Receive activity is aborted
All DMA activity is terminated
Cleared
Cleared
Halt operation
18.4.2
User Initialization (Prior to Asserting ECR[ETHER_EN])
The user needs to initialize portions the FEC prior to setting the ECR[ETHER_EN] bit. The exact values will depend on the particular application. The sequence is not important.
Ethernet MAC registers requiring initialization are defined in
Table 18-2. User Initialization (Before ECR[ETHER_EN])
Description
Initialize EIMR
Clear EIR (write 0xFFFF_FFFF)
TFWR (optional)
IALR / IAUR
GAUR / GALR
PALR / PAUR
OPD (only needed for full duplex flow control)
RCR
TCR
MSCR (optional)
Clear MIB_RAM (locations IPSBAR + 0x1200-0x12FC)
FEC FIFO/DMA registers that require initialization are defined in Table 18-3
.
Table 18-3. FEC User Initialization (Before ECR[ETHER_EN])
Description
Initialize FRSR (optional)
Initialize EMRBR
Initialize ERDSR
Initialize ETDSR
Initialize (Empty) Transmit Descriptor ring
Initialize (Empty) Receive Descriptor ring
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-5
Fast Ethernet Controller (FEC)
18.4.3
Microcontroller Initialization
In the FEC, the descriptor control RISC initializes some registers after ECR[ETHER_EN] is asserted.
After the microcontroller initialization sequence is complete, the hardware is ready for operation.
shows microcontroller initialization operations.
Table 18-4. Microcontroller Initialization
Description
Initialize BackOff Random Number Seed
Activate Receiver
Activate Transmitter
Clear Transmit FIFO
Clear Receive FIFO
Initialize Transmit Ring Pointer
Initialize Receive Ring Pointer
Initialize FIFO Count Registers
18.4.4
User Initialization (After Asserting ECR[ETHER_EN])
After asserting ECR[ETHER_EN], the user can set up the buffer/frame descriptors and write to the TDAR
and RDAR. Refer to Section 18.6, “Buffer Descriptors
” for more details.
18.4.5
Network Interface Options
The FEC supports both an MII interface for 10/100 Mbps Ethernet and a 7-wire serial interface for 10
Mbps Ethernet. The interface mode is selected by the RCR[MII_MODE] bit. In MII mode
(RCR[MII_MODE] = 1), there are 18 signals defined by the IEEE 802.3 standard and supported by the
EMAC. These signals are shown in Table 18-5 below.
Table 18-5. MII Mode
Signal Description
Transmit Clock
Transmit Enable
Transmit Data
Transmit Error
Collision
Carrier Sense
Receive Clock
Receive Data Valid
Receive Data
Receive Error
EMAC pin
ETXCLK
ETXEN
ETXD[3:0]
ETXER
ECOL
ECRS
ERXCLK
ERXDV
ERXD[3:0]
ERXER
18-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Functional Description
Table 18-5. MII Mode (continued)
Signal Description
Management Data Clock
Management Data
Input/Output
EMAC pin
EMDC
EMDIO
The 7-wire serial mode interface (RCR[MII_MODE] = 0) operates in what is generally referred to as the
“AMD” mode. 7-wire mode connections to the external transceiver are shown in Table 18-6
.
Table 18-6. 7-Wire Mode Configuration
SIGNAL DESCRIPTION
Transmit Clock
Transmit Enable
Transmit Data
Collision
Receive Clock
Receive Data Valid
Receive Data
EMAC PIN
ETXCLK
ETXEN
ETXD[0]
ECOL
ERXCLK
ERXDV
ERXD[0]
18.4.6
FEC Frame Transmission
The Ethernet transmitter is designed to work with almost no intervention from software. Once
ECR[ETHER_EN] is asserted and data appears in the transmit FIFO, the Ethernet MAC is able to transmit onto the network.
When the transmit FIFO fills to the watermark (defined by the TFWR), the MAC transmit logic will assert
ETXEN and start transmitting the preamble (PA) sequence, the start frame delimiter (SFD), and then the frame information from the FIFO. However, the controller defers the transmission if the network is busy
(ECRS asserts). Before transmitting, the controller waits for carrier sense to become inactive, then determines if carrier sense stays inactive for 60 bit times. If so, the transmission begins after waiting an additional 36 bit times (96 bit times after carrier sense originally became inactive). See
“Transmission Errors ” for more details.
If a collision occurs during transmission of the frame (half duplex mode), the Ethernet controller follows the specified backoff procedures and attempts to retransmit the frame until the retry limit is reached. The transmit FIFO stores at least the first 64 bytes of the transmit frame, so that they do not have to be retrieved from system memory in case of a collision. This improves bus utilization and latency in case immediate retransmission is necessary.
When all the frame data has been transmitted, the FCS (Frame Check Sequence or 32-bit Cyclic
Redundancy Check, CRC) bytes are appended if the TC bit is set in the transmit frame control word. If the
ABC bit is set in the transmit frame control word, a bad CRC will be appended to the frame data regardless of the TC bit value. Following the transmission of the CRC, the Ethernet controller writes the frame status information to the MIB block. Short frames are automatically padded by the transmit logic (if the TC bit in the transmit buffer descriptor for the end of frame buffer = 1).
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-7
Fast Ethernet Controller (FEC)
Both buffer (TXB) and frame (TFINT) interrupts may be generated as determined by the settings in the
EIMR.
The transmit error interrupts are HBERR, BABT, LATE_COL, COL_RETRY_LIM, and XFIFO_UN. If the transmit frame length exceeds MAX_FL bytes the BABT interrupt will be asserted, however the entire frame will be transmitted (no truncation).
To pause transmission, set the GTS (graceful transmit stop) bit in the TCR register. When the TCR[GTS] is set, the FEC transmitter stops immediately if transmission is not in progress; otherwise, it continues transmission until the current frame either finishes or terminates with a collision. After the transmitter has stopped the GRA (graceful stop complete) interrupt is asserted. If TCR[GTS] is cleared, the FEC resumes transmission with the next frame.
The Ethernet controller transmits bytes least significant bit first.
18.4.7
FEC Frame Reception
The FEC receiver is designed to work with almost no intervention from the host and can perform address recognition, CRC checking, short frame checking, and maximum frame length checking.
When the driver enables the FEC receiver by asserting ECR[ETHER_EN], it will immediately start processing receive frames. When ERXDV asserts, the receiver will first check for a valid PA/SFD header.
If the PA/SFD is valid, it will be stripped and the frame will be processed by the receiver. If a valid PA/SFD is not found, the frame will be ignored.
In serial mode, the first 16 bit times of RX_D0 following assertion of ERXDV are ignored. Following the first 16 bit times the data sequence is checked for alternating 1/0s. If a 11 or 00 data sequence is detected during bit times 17 to 21, the remainder of the frame is ignored. After bit time 21, the data sequence is monitored for a valid SFD (11). If a 00 is detected, the frame is rejected. When a 11 is detected, the PA/SFD sequence is complete.
In MII mode, the receiver checks for at least one byte matching the SFD. Zero or more PA bytes may occur, but if a 00 bit sequence is detected prior to the SFD byte, the frame is ignored.
After the first 6 bytes of the frame have been received, the FEC performs address recognition on the frame.
Once a collision window (64 bytes) of data has been received and if address recognition has not rejected the frame, the receive FIFO is signalled that the frame is “accepted” and may be passed on to the DMA.
If the frame is a runt (due to collision) or is rejected by address recognition, the receive FIFO is notified to “reject” the frame. Thus, no collision fragments are presented to the user except late collisions, which indicate serious LAN problems.
During reception, the Ethernet controller checks for various error conditions and once the entire frame is written into the FIFO, a 32-bit frame status word is written into the FIFO. This status word contains the
M, BC, MC, LG, NO, CR, OV and TR status bits, and the frame length. See
Receive Buffer (RXB) and Frame Interrupts (RFINT) may be generated if enabled by the EIMR register.
A receive error interrupt is babbling receiver error (BABR). Receive frames are not truncated if they exceed the max frame length (MAX_FL); however, the BABR interrupt will occur and the LG bit in the
When the receive frame is complete, the FEC sets the L-bit in the RxBD, writes the other frame status bits into the RxBD, and clears the E-bit. The Ethernet controller next generates a maskable interrupt (RFINT bit in EIR, maskable by RFIEN bit in EIMR), indicating that a frame has been received and is in memory.
The Ethernet controller then waits for a new frame.
18-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Functional Description
The Ethernet controller receives serial data LSB first.
18.4.8
Ethernet Address Recognition
The FEC filters the received frames based on destination address (DA) type — individual (unicast), group
(multicast), or broadcast (all-ones group address). The difference between an individual address and a group address is determined by the I/G bit in the destination address field. A flowchart for address recognition on received frames is illustrated in the figures below.
Address recognition is accomplished through the use of the receive block and microcode running on the microcontroller. The flowchart shown in
illustrates the address recognition decisions made by
the receive block, while Figure 18-3 illustrates the decisions made by the microcontroller.
If the DA is a broadcast address and broadcast reject (RCR[BC_REJ]) is deasserted, then the frame will
the microcontroller runs the address recognition subroutine, as shown in Figure 18-3 .
If the DA is a group (multicast) address and flow control is disabled, then the microcontroller will perform a group hash table lookup using the 64-entry hash table programmed in GAUR and GALR. If a hash match occurs, the receiver accepts the frame.
If flow control is enabled, the microcontroller will do an exact address match check between the DA and the designated PAUSE DA (01:80:C2:00:00:01). If the receive block determines that the received frame is a valid PAUSE frame, then the frame will be rejected. Note the receiver will detect a PAUSE frame with the DA field set to either the designated PAUSE DA or the unicast physical address.
If the DA is the individual (unicast) address, the microcontroller performs an individual exact match comparison between the DA and 48-bit physical address that the user programs in the PALR and PAUR registers. If an exact match occurs, the frame is accepted; otherwise, the microcontroller does an individual hash table lookup using the 64-entry hash table programmed in registers, IAUR and IALR. In the case of an individual hash match, the frame is accepted. Again, the receiver will accept or reject the frame based on PAUSE frame detection, shown in
.
If neither a hash match (group or individual), nor an exact match (group or individual) occur, then if promiscuous mode is enabled (RCR[PROM] = 1), then the frame will be accepted and the MISS bit in the receive buffer descriptor is set; otherwise, the frame will be rejected.
Similarly, if the DA is a broadcast address, broadcast reject (RCR[BC_REJ]) is asserted, and promiscuous mode is enabled, then the frame will be accepted and the MISS bit in the receive buffer descriptor is set; otherwise, the frame will be rejected.
In general, when a frame is rejected, it is flushed from the FIFO.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-9
Fast Ethernet Controller (FEC)
Accept/Reject
Frame
True
Broadcast Addr
?
False
Receive
Address
Recognition
False
Receive Frame
Set BC bit in RCV BD
BC_REJ = 1
?
True
False
Reject Frame
Flush from FIFO
PROM = 1
?
True
Hash Match
?
True
False
Exact Match
?
Receive Frame
Set MC bit in RCV BD if multicast
True
False
Pause Frame
?
False
True
Reject Frame
Flush from FIFO
Receive Frame Receive Frame
Set M (Miss) bit in Rcv BD
Set MC bit in Rcv BD if multicast
Set BC bit in Rcv BD if broadcast
NOTES:
BC_REJ - field in RCR register (BroadCast REJect)
PROM - field in RCR register (PROMiscous mode)
Pause Frame - valid PAUSE frame received
Figure 18-2. Ethernet Address Recognition—Receive Block Decisions
18-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Functional Description
Receive Address
Recognition
Group
I/G Address
?
FCE
?
False
False
True
Pause Address
?
True
Receive Frame
Hash Search
Group Table
Individual
False
Hash Search
Individual Table
Exact Match
?
True
Receive Frame
Match
?
False
True
Receive Frame
True
Match
?
False
Reject Frame
Flush from FIFO
Receive Frame
Reject Frame
Flush from FIFO
NOTES:
FCE - field in RCR register (Flow Control Enable)
I/G - Individual/Group bit in Destination Address (least significant bit in first byte received in MAC frame)
Figure 18-3. Ethernet Address Recognition—Microcode Decisions
18.4.9
Hash Algorithm
The hash table algorithm used in the group and individual hash filtering operates as follows. The 48-bit destination address is mapped into one of 64 bits, which are represented by 64 bits stored in GAUR, GALR
(group address hash match) or IAUR, IALR (individual address hash match). This mapping is performed by passing the 48-bit address through the on-chip 32-bit CRC generator and selecting the 6 most significant bits of the CRC-encoded result to generate a number between 0 and 63. The MSB of the CRC result selects GAUR (MSB = 1) or GALR (MSB = 0). The least significant 5 bits of the hash result select the bit within the selected register. If the CRC generator selects a bit that is set in the hash table, the frame is accepted; otherwise, it is rejected.
For example, if eight group addresses are stored in the hash table and random group addresses are received, the hash table prevents roughly 56/64 (or 87.5%) of the group address frames from reaching memory.
Those that do reach memory must be further filtered by the processor to determine if they truly contain one of the eight desired addresses.
The effectiveness of the hash table declines as the number of addresses increases.
The hash table registers must be initialized by the user. The CRC32 polynomial to use in computing the hash is:
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-11
Fast Ethernet Controller (FEC)
X 32 + X 26 + X 23 + X 22 + X 16 + X 12 + X 11 + X 10 + X 8 + X 7 + X 5 + X 4 + X 2 + X 1
A table of example Destination Addresses and corresponding hash values is included below for reference.
Table 18-7. Destination Address to 6-Bit Hash
48-bit DA
27:ff:ff:ff:ff:ff
07:ff:ff:ff:ff:ff
57:ff:ff:ff:ff:ff
77:ff:ff:ff:ff:ff f7:ff:ff:ff:ff:ff c7:ff:ff:ff:ff:ff
97:ff:ff:ff:ff:ff a7:ff:ff:ff:ff:ff
99:ff:ff:ff:ff:ff b9:ff:ff:ff:ff:ff f9:ff:ff:ff:ff:ff c9:ff:ff:ff:ff:ff
65:ff:ff:ff:ff:ff
55:ff:ff:ff:ff:ff
15:ff:ff:ff:ff:ff
35:ff:ff:ff:ff:ff b5:ff:ff:ff:ff:ff
95:ff:ff:ff:ff:ff d5:ff:ff:ff:ff:ff f5:ff:ff:ff:ff:ff db:ff:ff:ff:ff:ff fb:ff:ff:ff:ff:ff bb:ff:ff:ff:ff:ff
8b:ff:ff:ff:ff:ff
0b:ff:ff:ff:ff:ff
3b:ff:ff:ff:ff:ff
7b:ff:ff:ff:ff:ff
5b:ff:ff:ff:ff:ff
Hash Decimal
Value
20
21
22
23
16
17
18
19
24
25
26
27
12
13
14
15
10
11
8
9
6
7
4
5
2
3
0
1
6-bit Hash (in hex)
0x10
0x11
0x12
0x13
0x14
0x15
0x16
0x17
0x18
0x19
0x1a
0x1b
0xc
0xd
0xe
0xf
0x8
0x9
0xa
0xb
0x4
0x5
0x6
0x7
0x0
0x1
0x2
0x3
18-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
bf:ff:ff:ff:ff:ff
9f:ff:ff:ff:ff:ff df:ff:ff:ff:ff:ff ef:ff:ff:ff:ff:ff
93:ff:ff:ff:ff:ff b3:ff:ff:ff:ff:ff f3:ff:ff:ff:ff:ff d3:ff:ff:ff:ff:ff
53:ff:ff:ff:ff:ff
73:ff:ff:ff:ff:ff
23:ff:ff:ff:ff:ff
13:ff:ff:ff:ff:ff
3d:ff:ff:ff:ff:ff
0d:ff:ff:ff:ff:ff
5d:ff:ff:ff:ff:ff
7d:ff:ff:ff:ff:ff
59:ff:ff:ff:ff:ff
79:ff:ff:ff:ff:ff
29:ff:ff:ff:ff:ff
19:ff:ff:ff:ff:ff d1:ff:ff:ff:ff:ff f1:ff:ff:ff:ff:ff b1:ff:ff:ff:ff:ff
91:ff:ff:ff:ff:ff
11:ff:ff:ff:ff:ff
31:ff:ff:ff:ff:ff
71:ff:ff:ff:ff:ff
51:ff:ff:ff:ff:ff
7f:ff:ff:ff:ff:ff
4f:ff:ff:ff:ff:ff
1f:ff:ff:ff:ff:ff
3f:ff:ff:ff:ff:ff
Table 18-7. Destination Address to 6-Bit Hash (continued)
48-bit DA
6-bit Hash (in hex)
0x34
0x35
0x36
0x37
0x38
0x39
0x3a
0x3b
0x2c
0x2d
0x2e
0x2f
0x30
0x31
0x32
0x33
0x24
0x25
0x26
0x27
0x28
0x29
0x2a
0x2b
0x1c
0x1d
0x1e
0x1f
0x20
0x21
0x22
0x23
Hash Decimal
Value
56
57
58
59
52
53
54
55
48
49
50
51
44
45
46
47
40
41
42
43
36
37
38
39
32
33
34
35
28
29
30
31
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary
Functional Description
18-13
Fast Ethernet Controller (FEC)
Table 18-7. Destination Address to 6-Bit Hash (continued)
48-bit DA fd:ff:ff:ff:ff:ff dd:ff:ff:ff:ff:ff
9d:ff:ff:ff:ff:ff bd:ff:ff:ff:ff:ff
6-bit Hash (in hex)
0x3c
0x3d
0x3e
0x3f
Hash Decimal
Value
60
61
62
63
18.4.10 Full Duplex Flow Control
Full-duplex flow control allows the user to transmit pause frames and to detect received pause frames.
Upon detection of a pause frame, MAC data frame transmission stops for a given pause duration.
To enable pause frame detection, the FEC must operate in full-duplex mode (TCR[FDEN] asserted) and flow control enable (RCR[FCE]) must be asserted. The FEC detects a pause frame when the fields of the incoming frame match the pause frame specifications, as shown in the table below. In addition, the receive status associated with the frame should indicate that the frame is valid.
Table 18-8. PAUSE Frame Field Specification
48-bit Destination Address
48-bit Source Address
16-bit Type
16-bit Opcode
16-bit PAUSE Duration
0x0180_c200_0001 or Physical Address
Any
0x8808
0x0001
0x0000 to 0xFFFF
Pause frame detection is performed by the receiver and microcontroller modules. The microcontroller runs an address recognition subroutine to detect the specified pause frame destination address, while the receiver detects the type and opcode pause frame fields. On detection of a pause frame, TCR[GTS] is asserted by the FEC internally. When transmission has paused, the EIR[GRA] interrupt is asserted and the pause timer begins to increment. Note that the pause timer makes use of the transmit backoff timer hardware, which is used for tracking the appropriate collision backoff time in half-duplex mode. The pause timer increments once every slot time, until OPD[PAUSE_DUR] slot times have expired. On
OPD[PAUSE_DUR] expiration, TCR[GTS] is deasserted allowing MAC data frame transmission to resume. Note that the receive flow control pause (TCR[RFC_PAUSE]) status bit is asserted while the transmitter is paused due to reception of a pause frame.
To transmit a pause frame, the FEC must operate in full-duplex mode and the user must assert flow control pause (TCR[TFC_PAUSE]). On assertion of transmit flow control pause (TCR[TFC_PAUSE]), the transmitter asserts TCR[GTS] internally. When the transmission of data frames stops, the EIR[GRA]
(graceful stop complete) interrupt asserts. Following EIR[GRA] assertion, the pause frame is transmitted.
On completion of pause frame transmission, flow control pause (TCR[TFC_PAUSE]) and TCR[GTS] are deasserted internally.
The user must specify the desired pause duration in the OPD register.
Note that when the transmitter is paused due to receiver/microcontroller pause frame detection, transmit flow control pause (TCR[TFC_PAUSE]) still may be asserted and will cause the transmission of a single pause frame. In this case, the EIR[GRA] interrupt will not be asserted.
18-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Functional Description
18.4.11 Inter-Packet Gap (IPG) Time
The minimum inter-packet gap time for back-to-back transmission is 96 bit times. After completing a transmission or after the backoff algorithm completes, the transmitter waits for carrier sense to be negated before starting its 96 bit time IPG counter. Frame transmission may begin 96 bit times after carrier sense is negated if it stays negated for at least 60 bit times. If carrier sense asserts during the last 36 bit times, it will be ignored and a collision will occur.
The receiver receives back-to-back frames with a minimum spacing of at least 28 bit times. If an inter-packet gap between receive frames is less than 28 bit times, the following frame may be discarded by the receiver.
18.4.12 Collision Handling
If a collision occurs during frame transmission, the Ethernet controller will continue the transmission for at least 32 bit times, transmitting a JAM pattern consisting of 32 ones. If the collision occurs during the preamble sequence, the JAM pattern will be sent after the end of the preamble sequence.
If a collision occurs within 512 bit times, the retry process is initiated. The transmitter waits a random number of slot times. A slot time is 512 bit times. If a collision occurs after 512 bit times, then no retransmission is performed and the end of frame buffer is closed with a Late Collision (LC) error indication.
18.4.13 Internal and External Loopback
Both internal and external loopback are supported by the Ethernet controller. In loopback mode, both of the FIFOs are used and the FEC actually operates in a full-duplex fashion. Both internal and external loopback are configured using combinations of the LOOP and DRT bits in the RCR register and the FDEN bit in the TCR register.
For both internal and external loopback set FDEN = 1.
For internal loopback set RCR[LOOP] = 1 and RCR[DRT] = 0. ETXEN and ETXER will not assert during internal loopback. During internal loopback, the transmit/receive data rate is higher than in normal operation because the internal system clock is used by the transmit and receive blocks instead of the clocks from the external transceiver. This will cause an increase in the required system bus bandwidth for transmit and receive data being DMA’d to/from external memory. It may be necessary to pace the frames on the transmit side and/or limit the size of the frames to prevent transmit FIFO underrun and receive FIFO overflow.
For external loopback set RCR[LOOP] = 0, RCR[DRT] = 0 and configure the external transceiver for loopback.
18.4.14 Ethernet Error-Handling Procedure
The Ethernet controller reports frame reception and transmission error conditions using the FEC RxBDs, the EIR register, and the MIB block counters.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-15
Fast Ethernet Controller (FEC)
18.4.14.1 Transmission Errors
18.4.14.1.1 Transmitter Underrun
If this error occurs, the FEC sends 32 bits that ensure a CRC error and stops transmitting. All remaining buffers for that frame are then flushed and closed. The UN bit is set in the EIR. The FEC will then continue to the next transmit buffer descriptor and begin transmitting the next frame.
The “UN” interrupt will be asserted if enabled in the EIMR register.
18.4.14.1.2 Retransmission Attempts Limit Expired
When this error occurs, the FEC terminates transmission. All remaining buffers for that frame are flushed and closed, and the RL bit is set in the EIR. The FEC will then continue to the next transmit buffer descriptor and begin transmitting the next frame.
The “RL” interrupt will be asserted if enabled in the EIMR register.
18.4.14.1.3 Late Collision
When a collision occurs after the slot time (512 bits starting at the Preamble), the FEC terminates transmission. All remaining buffers for that frame are flushed and closed, and the LC bit is set in the EIR register. The FEC will then continue to the next transmit buffer descriptor and begin transmitting the next frame.
The “LC” interrupt will be asserted if enabled in the EIMR register.
18.4.14.1.4 Heartbeat
Some transceivers have a self-test feature called “heartbeat” or “signal quality error.” To signify a good self-test, the transceiver indicates a collision to the FEC within 4 microseconds after completion of a frame transmitted by the Ethernet controller. This indication of a collision does not imply a real collision error on the network, but is rather an indication that the transceiver still seems to be functioning properly. This is called the heartbeat condition.
If the HBC bit is set in the TCR register and the heartbeat condition is not detected by the FEC after a frame transmission, then a heartbeat error occurs. When this error occurs, the FEC closes the buffer, sets the HB bit in the EIR register, and generates the HBERR interrupt if it is enabled.
18.4.14.2 Reception Errors
18.4.14.2.1 Overrun Error
If the receive block has data to put into the receive FIFO and the receive FIFO is full, the FEC sets the OV bit in the RxBD. All subsequent data in the frame will be discarded and subsequent frames may also be discarded until the receive FIFO is serviced by the DMA and space is made available. At this point the receive frame/status word is written into the FIFO with the OV bit set. This frame must be discarded by the driver.
18.4.14.2.2 Non-Octet Error (Dribbling Bits)
The Ethernet controller handles up to seven dribbling bits when the receive frame terminates past an non-octet aligned boundary. Dribbling bits are not used in the CRC calculation. If there is a CRC error,
18-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Programming Model then the frame non-octet aligned (NO) error is reported in the RxBD. If there is no CRC error, then no error is reported.
18.4.14.2.3 CRC Error
When a CRC error occurs with no dribble bits, the FEC closes the buffer and sets the CR bit in the RxBD.
CRC checking cannot be disabled, but the CRC error can be ignored if checking is not required.
18.4.14.2.4 Frame Length Violation
When the receive frame length exceeds MAX_FL bytes the BABR interrupt will be generated, and the LG bit in the end of frame RxBD will be set. The frame is not truncated unless the frame length exceeds 2047 bytes).
18.4.14.2.5 Truncation
When the receive frame length exceeds 2047 bytes the frame is truncated and the TR bit is set in the receive
BD.
18.5
Programming Model
This section gives an overview of the registers, followed by a description of the buffers.
The FEC is programmed by a combination of control/status registers (CSRs) and buffer descriptors. The
CSRs are used for mode control and to extract global status information. The descriptors are used to pass data buffers and related buffer information between the hardware and software.
18.5.1
Top Level Module Memory Map
The FEC implementation requires a 1-Kbyte memory map space. This is divided into 2 sections of 512 bytes each. The first is used for control/status registers. The second contains event/statistic counters held
in the MIB block. Table 18-9 defines the top level memory map.
Table 18-9. Module Memory Map
Address
IPSBAR + 0x1000-11FF
IPSBAR + 0x1200-13FF
Function
Control/Status Registers
MIB Block Counters
18.5.2
Detailed Memory Map (Control/Status Registers)
shows the FEC register memory map with each register address, name, and a brief description.
Table 18-10. FEC Register Memory Map
IPSBAR
Offset
0x1004
0x1008
0x1010
0x1014
Name
EIR
EIMR
RDAR
TDAR
Width
32
32
32
32
Description
Interrupt Event Register
Interrupt Mask Register
Receive Descriptor Active Register
Transmit Descriptor Active Register
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-17
Fast Ethernet Controller (FEC)
0x1120
0x1124
0x1144
0x114C
0x1150
0x1180
0x1184
0x1188
IPSBAR
Offset
0x1024
0x1040
0x1044
0x1064
0x1084
0x10C4
0x10E4
0x10E8
0x10EC
0x1118
0x111C
Table 18-10. FEC Register Memory Map (continued)
GAUR
GALR
TFWR
FRBR
FRSR
ERDSR
ETDSR
EMRBR
Name
ECR
MDATA
MSCR
MIBC
RCR
TCR
PALR
PAUR
OPD
IAUR
IALR
Width Description
32
32
32
32
32
32
32
32
32
32
32
32
Ethernet Control Register
MII Data Register
MII Speed Control Register
MIB Control/Status Register
Receive Control Register
Transmit Control Register
Physical Address Low Register
Physical Address High+ Type Field
32
32
32
32
32 Opcode + Pause Duration
32 Upper 32 bits of Individual Hash Table
32 Lower 32 Bits of Individual Hash
Table
Upper 32 bits of Group Hash Table
Lower 32 bits of Group Hash Table
Transmit FIFO Watermark
FIFO Receive Bound Register
FIFO Receive FIFO Start Registers
Pointer to Receive Descriptor Ring
Pointer to Transmit Descriptor Ring
Maximum Receive Buffer Size
18.5.3
MIB Block Counters Memory Map
Table 18-11 defines the MIB Counters memory map which defines the locations in the MIB RAM space
where hardware maintained counters reside. These fall in the 0x1200-0x13FF address offset range. The counters are divided into two groups.
RMON counters are included which cover the Ethernet Statistics counters defined in RFC 1757. In addition to the counters defined in the Ethernet Statistics group, a counter is included to count truncated frames as the FEC only supports frame lengths up to 2047 bytes. The RMON counters are implemented independently for transmit and receive to insure accurate network statistics when operating in full duplex mode.
IEEE counters are included which support the Mandatory and Recommended counter packages defined in section 5 of ANSI/IEEE Std. 802.3 (1998 edition). The IEEE Basic Package objects are supported by the
FEC but do not require counters in the MIB block. In addition, some of the recommended package objects which are supported do not require MIB counters. Counters for transmit and receive full duplex flow control frames are included as well.
18-18
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
IPSBAR
Offset
0x1260
0x1264
0x1268
0x126C
0x1270
0x1274
0x1284
0x1288
0x1240
0x1244
0x1248
0x124C
0x1250
0x1254
0x1258
0x125c
0x1220
0x1224
0x1228
0x122C
0x1230
0x1234
0x1238
0x123C
0x1200
0x1204
0x1208
0x120C
0x1210
0x1214
0x1218
0x121C
Programming Model
Table 18-11. MIB Counters Memory Map
Mnemonic
RMON_T_DROP
RMON_T_PACKETS
RMON_T_BC_PKT
RMON_T_MC_PKT
RMON_T_CRC_ALIGN
RMON_T_UNDERSIZE
RMON_T_OVERSIZE
RMON_T_FRAG
RMON_T_JAB
RMON_T_COL
RMON_T_P64
RMON_T_P65TO127
RMON_T_P128TO255
RMON_T_P256TO511
RMON_T_P512TO1023
RMON_T_P1024TO2047
RMON_T_P_GTE2048
RMON_T_OCTETS
IEEE_T_DROP
IEEE_T_FRAME_OK
IEEE_T_1COL
IEEE_T_MCOL
IEEE_T_DEF
IEEE_T_LCOL
IEEE_T_EXCOL
IEEE_T_MACERR
IEEE_T_CSERR
IEEE_T_SQE
IEEE_T_FDXFC
IEEE_T_OCTETS_OK
RMON_R_PACKETS
RMON_R_BC_PKT
Description
Count of frames not counted correctly
RMON Tx packet count
RMON Tx Broadcast Packets
RMON Tx Multicast Packets
RMON Tx Packets w CRC/Align error
RMON Tx Packets < 64 bytes, good crc
RMON Tx Packets > MAX_FL bytes, good crc
RMON Tx Packets < 64 bytes, bad crc
RMON Tx Packets > MAX_FL bytes, bad crc
RMON Tx collision count
RMON Tx 64 byte packets
RMON Tx 65 to 127 byte packets
RMON Tx 128 to 255 byte packets
RMON Tx 256 to 511 byte packets
RMON Tx 512 to 1023 byte packets
RMON Tx 1024 to 2047 byte packets
RMON Tx packets w > 2048 bytes
RMON Tx Octets
Count of frames not counted correctly
Frames Transmitted OK
Frames Transmitted with Single Collision
Frames Transmitted with Multiple Collisions
Frames Transmitted after Deferral Delay
Frames Transmitted with Late Collision
Frames Transmitted with Excessive Collisions
Frames Transmitted with Tx FIFO Underrun
Frames Transmitted with Carrier Sense Error
Frames Transmitted with SQE Error
Flow Control Pause frames transmitted
Octet count for Frames Transmitted w/o Error
RMON Rx packet count
RMON Rx Broadcast Packets
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-19
Fast Ethernet Controller (FEC)
IPSBAR
Offset
0x12AC
0x12B0
0x12B4
0x12B8
0x12BC
0x12C0
0x12C4
0x12C8
0x128C
0x1290
0x1294
0x1298
0x129C
0x12A0
0x12A4
0x12A8
0x12CC
0x12D0
0x12D4
0x12D8
0x12DC
0x12E0
Table 18-11. MIB Counters Memory Map (continued)
Mnemonic
RMON_R_MC_PKT
RMON_R_CRC_ALIGN
RMON_R_UNDERSIZE
RMON_R_OVERSIZE
RMON_R_FRAG
RMON_R_JAB
RMON_R_RESVD_0
RMON_R_P64
RMON_R_P65TO127
RMON_R_P128TO255
RMON_R_P256TO511
RMON_R_P512TO1023
RMON_R_P1024TO2047
RMON_R_P_GTE2048
RMON_R_OCTETS
IEEE_R_DROP
IEEE_R_FRAME_OK
IEEE_R_CRC
IEEE_R_ALIGN
IEEE_R_MACERR
IEEE_R_FDXFC
IEEE_R_OCTETS_OK
Description
RMON Rx Multicast Packets
RMON Rx Packets w CRC/Align error
RMON Rx Packets < 64 bytes, good crc
RMON Rx Packets > MAX_FL bytes, good crc
RMON Rx Packets < 64 bytes, bad crc
RMON Rx Packets > MAX_FL bytes, bad crc
RMON Rx 64 byte packets
RMON Rx 65 to 127 byte packets
RMON Rx 128 to 255 byte packets
RMON Rx 256 to 511 byte packets
RMON Rx 512 to 1023 byte packets
RMON Rx 1024 to 2047 byte packets
RMON Rx packets w > 2048 bytes
RMON Rx Octets
Count of frames not counted correctly
Frames Received OK
Frames Received with CRC Error
Frames Received with Alignment Error
Receive Fifo Overflow count
Flow Control Pause frames received
Octet count for Frames Rcvd w/o Error
18.5.4
Registers
The following sections describe each register in detail.
18.5.4.1
Ethernet Interrupt Event Register (EIR)
When an event occurs that sets a bit in the EIR, an interrupt will be generated if the corresponding bit in the interrupt mask register (EIMR) is also set. The bit in the EIR is cleared if a one is written to that bit position; writing zero has no effect. This register is cleared upon hardware reset.
These interrupts can be divided into operational interrupts, transceiver/network error interrupts, and internal error interrupts. Interrupts which may occur in normal operation are GRA, TXF, TXB, RXF, RXB, and MII. Interrupts resulting from errors/problems detected in the network or transceiver are HBERR,
BABR, BABT, LC and RL. Interrupts resulting from internal errors are HBERR and UN.
18-20
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Programming Model
Some of the error interrupts are independently counted in the MIB block counters. Software may choose to mask off these interrupts since these errors will be visible to network management via the MIB counters.
• HBERR - IEEE_T_SQE
• BABR - RMON_R_OVERSIZE (good CRC), RMON_R_JAB (bad CRC)
• BABT - RMON_T_OVERSIZE (good CRC), RMON_T_JAB (bad CRC)
• LATE_COL - IEEE_T_LCOL
• COL_RETRY_LIM - IEEE_T_EXCOL
• XFIFO_UN - IEEE_T_MACERR
31 30 29 28 27 26 25 24 23 22 21 20 19
Field HBERR BABR BABT GRA TXF TXB RXF RXB MII EBERR LC RL UN
Reset 0000_0000_0000_0000
R/W R/W
18
—
16
15 0
Field
Reset
R/W
Address
—
0000_0000_0000_0000
R/W
IPSBAR + 0x1004
Figure 18-4. Ethernet Interrupt Event Register (EIR)
Bits
31
30
29
28
Name
HBERR
BABR
BABT
GRA
Table 18-12. EIR Field Descriptions
Description
Heartbeat error. This interrupt indicates that HBC is set in the TCR register and that the COL input was not asserted within the
Heartbeat window following a transmission.
Babbling receive error. This bit indicates a frame was received with length in excess of RCR[MAX_FL] bytes.
Babbling transmit error. This bit indicates that the transmitted frame length has exceeded RCR[MAX_FL] bytes. This condition is usually caused by a frame that is too long being placed into the transmit data buffer(s). Truncation does not occur.
Graceful stop complete. This interrupt will be asserted for one of three reasons. Graceful stop means that the transmitter is put into a pause state after completion of the frame currently being transmitted.
1) A graceful stop, which was initiated by the setting of the
TCR[GTS] bit is now complete.
2) A graceful stop, which was initiated by the setting of the
TCR[TFC_PAUSE] bit is now complete.
3) A graceful stop, which was initiated by the reception of a valid full duplex flow control “pause” frame is now complete. Refer to the “Full Duplex Flow Control” section of the Functional
Description chapter.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-21
Fast Ethernet Controller (FEC)
Bits
27
26
25
24
23
22
21
20
19
18–0
Table 18-12. EIR Field Descriptions (continued)
Name
TXF
TXB
RXF
RXB
MII
EBERR
LC
RL
UN
—
Description
Transmit frame interrupt. This bit indicates that a frame has been transmitted and that the last corresponding buffer descriptor has been updated.
Transmit buffer interrupt. This bit indicates that a transmit buffer descriptor has been updated.
Receive frame interrupt. This bit indicates that a frame has been received and that the last corresponding buffer descriptor has been updated.
Receive buffer interrupt. This bit indicates that a receive buffer descriptor has been updated that was not the last in the frame.
MII interrupt. This bit indicates that the MII has completed the data transfer requested.
Ethernet bus error. This bit indicates that a system bus error occurred when a DMA transaction was underway. When the
EBERR bit is set, ECR[ETHER_EN] will be cleared, halting frame processing by the FEC. When this occurs software will need to insure that the FIFO controller and DMA are also soft reset.
Late collison. This bit indicates that a collision occurred beyond the collision window (slot time) in half duplex mode. The frame is truncated with a bad CRC and the remainder of the frame is discarded.
Collision retry limit. This bit indicates that a collision occurred on each of 16 successive attempts to transmit the frame. The frame is discarded without being transmitted and transmission of the next frame will commence. Can only occur in half duplex mode.
Transmit FIFO underrun. This bit indicates that the transmit FIFO became empty before the complete frame was transmitted. A bad
CRC is appended to the frame fragment and the remainder of the frame is discarded.
Reserved, should be cleared.
18-22
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Programming Model
18.5.4.2
Interrupt Mask Register (EIMR)
The EIMR register controls which interrupt events are allowed to generate actual interrupts. All implemented bits in this CSR are read/write. This register is cleared upon a hardware reset. If the corresponding bits in both the EIR and EIMR registers are set, the interrupt will be signalled to the CPU.
The interrupt signal will remain asserted until a 1 is written to the EIR bit (write 1 to clear) or a 0 is written to the EIMR bit.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 16
Field HBERR BABR BABT GRA TXF TXB RXF RXB MII EBERR LC RL UN —
Reset 0000_0000_0000_0000
R/W R/W
15
Field
Reset
R/W
Address
Bits
31–19
18–0
—
0000_0000_0000_0000
R/W
IPSBAR + 0x1008
Figure 18-5. Interrupt Mask Register (EIMR)
Table 18-13. EIMR Field Descriptions
Name Description
See
.
Interrupt mask. Each bit corresponds to an interrupt source defined by the EIR register. The corresponding EIMR bit determines whether an interrupt condition can generate an interrupt. At every processor clock, the EIR samples the signal generated by the interrupting source. The corresponding EIR bit reflects the state of the interrupt signal even if the corresponding
EIMR bit is set.
0 The corresponding interrupt source is masked.
1 The corresponding interrupt source is not masked.
— Reserved, should be cleared.
0
18.5.4.3
Receive Descriptor Active Register (RDAR)
RDAR is a command register, written by the user, that indicates that the receive descriptor ring has been updated (empty receive buffers have been produced by the driver with the empty bit set).
Whenever the register is written, the RDAR bit is set. This is independent of the data actually written by the user. When set, the FEC will poll the receive descriptor ring and process receive frames (provided
ECR[ETHER_EN] is also set). Once the FEC polls a receive descriptor whose empty bit is not set, then the FEC will clear the RDAR bit and cease receive descriptor ring polling until the user sets the bit again, signifying that additional descriptors have been placed into the receive descriptor ring.
The RDAR register is cleared at reset and when ECR[ETHER_EN] is cleared.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-23
Fast Ethernet Controller (FEC)
Field
Reset
R/W
Field
Reset
R/W
Address
31
15
Bits
31–25
24
—
25 24
R_DES_ACTIVE
0000_0000_0000_0000
R/W
23
—
16
—
0000_0000_0000_0000
R/W
IPSBAR + 0x1010
Figure 18-6. Receive Descriptor Active Register (RDAR)
Table 18-14. RDAR Field Descriptions
Name Description
— Reserved, should be cleared.
R_DES_ACTIVE Set to one when this register is written, regardless of the value written.
Cleared by the FEC device whenever no additional “empty” descriptors remain in the receive ring. Also cleared when ECR[ETHER_EN] is cleared.
— Reserved, should be cleared.
0
23–0
18.5.4.4
Transmit Descriptor Active Register (TDAR)
The TDAR is a command register which should be written by the user to indicate that the transmit descriptor ring has been updated (transmit buffers have been produced by the driver with the ready bit set in the buffer descriptor).
Whenever the register is written, the TDAR bit is set. This value is independent of the data actually written by the user. When set, the FEC will poll the transmit descriptor ring and process transmit frames (provided
ECR[ETHER_EN] is also set). Once the FEC polls a transmit descriptor whose ready bit is not set, then the FEC will clear the TDAR bit and cease transmit descriptor ring polling until the user sets the bit again, signifying additional descriptors have been placed into the transmit descriptor ring.
The TDAR register is cleared at reset, when ECR[ETHER_EN] is cleared, or when ECR[RESET] is set.
18-24
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Programming Model
Field
Reset
R/W
31
—
25 24
X_DES_ACTIVE
0000_0000_0000_0000
R/W
23
—
16
15 0
Field
Reset
R/W
Address
Bits
31–25
24
—
0000_0000_0000_0000
R/W
IPSBAR + 0x1014
Figure 18-7. Transmit Descriptor Active Register (TDAR)
Table 18-15. TDAR Field Descriptions
Name Description
— Reserved, should be cleared.
X_DES_ACTIVE Set to one when this register is written, regardless of the value written.
Cleared by the FEC device whenever no additional “ready” descriptors remain in the transmit ring. Also cleared when ECR[ETHER_EN] is cleared.
— Reserved, should be cleared.
23–0
18.5.4.5
Ethernet Control Register (ECR)
ECR is a read/write user register, though both fields in this register may be altered by hardware as well.
The ECR is used to enable/disable the FEC.
16
Field
Reset
R/W
31
Field
Reset
R/W
Address
15
—
1111_0000_0000_0000
R/W
—
0000_0000_0000_0000
R/W
IPSBAR + 0x1024
Figure 18-8. Ethernet Control Register (ECR)
1 0
ETHER_EN RESET
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-25
Fast Ethernet Controller (FEC)
Bits
31-2
1
0
Table 18-16. ECR Field Descriptions
Name Description
— Reserved.
ETHER_EN When this bit is set, the FEC is enabled, and reception and transmission are possible. When this bit is cleared, reception is immediately stopped and transmission is stopped after a bad CRC is appended to any currently transmitted frame. The buffer descriptor(s) for an aborted transmit frame are not updated after clearing this bit. When ETHER_EN is deasserted, the DMA, buffer descriptor, and FIFO control logic are reset, including the buffer descriptor and FIFO pointers. The ETHER_EN bit is altered by hardware under the following conditions:
• ECR[RESET] is set by software, in which case ETHER_EN will be cleared
• An error condition causes the EIR[EBERR] bit to set, in which case ETHER_EN will be cleared
RESET When this bit is set, the equivalent of a hardware reset is performed but it is local to the FEC. ETHER_EN is cleared and all other FEC registers take their reset values. Also, any transmission/reception currently in progress is abruptly aborted.
This bit is automatically cleared by hardware during the reset sequence. The reset sequence takes approximately 8 system clock cycles after RESET is written with a 1.
18.5.4.6
MII Management Frame Register (MMFR)
The MMFR is accessed by the user and does not reset to a defined value. The MMFR register is used to communicate with the attached MII compatible PHY device(s), providing read/write access to their MII registers. Performing a write to the MMFR will cause a management frame to be sourced unless the MSCR has been programmed to 0. In the case of writing to MMFR when MSCR = 0, if the MSCR register is then written to a non-zero value, an MII frame will be generated with the data previously written to the MMFR.
This allows MMFR and MSCR to be programmed in either order if MSCR is currently zero.
RA
18 17 16
TA Field
Reset
R/W
31
ST
30
15
Field
Reset
R/W
Address
29
OP
28 27
PA
Undefined
R/W
23 22
DATA
Undefined
R/W
IPSBAR + 0x1040
Figure 18-9. MII Management Frame Register (MMFR)
0
18-26
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Programming Model
Bit
31–30
29–28
27–23
22–18
17–16
15–0
Table 18-17. MMFR Field Descriptions
Name
ST
OP
PA
RA
TA
DATA
Description
Start of frame delimiter. These bits must be programmed to 01 for a valid MII management frame.
Operation code. This field must be programmed to 10 (read) or
01 (write) to generate a valid MII management frame. A value of 11 will produce “read” frame operation while a value of 00 will produce “write” frame operation, but these frames will not be
MII compliant.
PHY address. This field specifies one of up to 32 attached PHY devices.
Register address. This field specifies one of up to 32 registers within the specified PHY device.
Turn around. This field must be programmed to 10 to generate a valid MII management frame.
Management frame data. This is the field for data to be written to or read from the PHY register.
To perform a read or write operation on the MII Management Interface, the MMFR register must be written by the user. To generate a valid read or write management frame, the ST field must be written with a 01 pattern, and the TA field must be written with a 10. If other patterns are written to these fields, a frame will be generated but will not comply with the IEEE 802.3 MII definition.
To generate an IEEE 802.3-compliant MII Management Interface write frame (write to a PHY register), the user must write {01 01 PHYAD REGAD 10 DATA} to the MMFR register. Writing this pattern will cause the control logic to shift out the data in the MMFR register following a preamble generated by the control state machine. During this time the contents of the MMFR register will be altered as the contents are serially shifted and will be unpredictable if read by the user. Once the write management frame operation has completed, the MII interrupt will be generated. At this time the contents of the MMFR register will match the original value written.
To generate an MII Management Interface read frame (read a PHY register) the user must write {01 10
PHYAD REGAD 10 XXXX} to the MMFR register (the content of the DATA field is a don’t care). Writing this pattern will cause the control logic to shift out the data in the MMFR register following a preamble generated by the control state machine. During this time the contents of the MMFR register will be altered as the contents are serially shifted, and will be unpredictable if read by the user. Once the read management frame operation has completed, the MII interrupt will be generated. At this time the contents of the MMFR register will match the original value written except for the DATA field whose contents have been replaced by the value read from the PHY register.
If the MMFR register is written while frame generation is in progress, the frame contents will be altered.
Software should use the MII_STATUS register and/or the MII interrupt to avoid writing to the MMFR register while frame generation is in progress.
18.5.4.7
MII Speed Control Register (MSCR)
The MSCR provides control of the MII clock (EMDC pin) frequency, allows a preamble drop on the MII management frame, and provides observability (intended for manufacturing test) of an internal counter used in generating the EMDC clock signal.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-27
Fast Ethernet Controller (FEC)
31 16
Field
Reset
R/W
Field
Reset
R/W
Address
15
—
0000_0000_0000_0000
R/W
—
8 7
DIS_PREAMBLE
6
0000_0000_0000_0000
R/W
IPSBAR + 0x1044
Figure 18-10. MII Speed Control Register (MSCR)
MII_SPEED
1 0
—
Bits
31–8
7
6–1
Table 18-18. MSCR Field Descriptions
Name Description
— Reserved, should be cleared.
DIS_PREAMBLE Asserting this bit will cause preamble (32 1’s) not to be prepended to the MII management frame. The MII standard allows the preamble to be dropped if the attached PHY device(s) does not require it.
MII_SPEED MII_SPEED controls the frequency of the MII management interface clock (EMDC) relative to the system clock. A value of 0 in this field will “turn off” the EMDC and leave it in low voltage state. Any non-zero value will result in the EMDC frequency of
1/(MII_SPEED*2) of the system clock frequency.
— Reserved, should be cleared.
0
The MII_SPEED field must be programmed with a value to provide an EMDC frequency of less than or equal to 2.5 MHz to be compliant with the IEEE 802.3 MII specification. The MII_SPEED must be set to a non-zero value in order to source a read or write management frame. After the management frame is complete the MSCR register may optionally be set to zero to turn off the EMDC. The EMDC generated will have a 50% duty cycle except when MII_SPEED is changed during operation (change will take effect following either a rising or falling edge of EMDC).
If the system clock is 25 MHz, programming this register to 0x0000_0005 will result in an EMDC frequency of 25 MHz * 1/10 = 2.5 MHz. A table showing optimum values for MII_SPEED as a function of system clock frequency is provided below.
Table 18-19. Programming Examples for MSCR
System Clock Frequency
25 MHz
33 MHz
40 MHz
MII_SPEED (field in reg)
0x5
0x7
0x8
EMDC frequency
2.5 MHz
2.36 MHz
2.5 MHz
18-28
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Programming Model
Table 18-19. Programming Examples for MSCR (continued)
System Clock Frequency
50 MHz
66 MHz
MII_SPEED (field in reg)
0xA
0xD
EMDC frequency
2.5 MHz
2.5 MHz
18.5.4.8
MIB Control Register (MIBC)
The MIBC is a read/write register used to provide control of and to observe the state of the MIB block.
This register is accessed by user software if there is a need to disable the MIB block operation. For example, in order to clear all MIB counters in RAM the user should disable the MIB block, then clear all the MIB RAM locations, then enable the MIB block. The MIB_DISABLE bit is reset to 1. See
for the locations of the MIB counters.
16 31 30
Field MIB_DISABLE MIB_IDLE
Reset
R/W
15
Field
Reset
R/W
Address
1100_0000_000_000
R/W
—
—
0000_0000_0000_0000
R/W
IPSBAR + 0x1064
Figure 18-11. MIB Control Register (MIBC)
Bits
31
30
29–0
Table 18-20. MIBC Field Descriptions
Name Description
MIB_DISABLE A read/write control bit. If set, the MIB logic will halt and not update any MIB counters.
MIB_IDLE
—
A read-only status bit. If set the MIB block is not currently updating any MIB counters.
Reserved.
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-29
Fast Ethernet Controller (FEC)
18.5.4.9
Receive Control Register (RCR)
The RCR is programmed by the user. The RCR controls the operational mode of the receive block and should be written only when ECR[ETHER_EN] = 0 (initialization time).
Field
Reset
R/W
31
Field
Reset
R/W
Address
15
Bits
31–27
26–16
15–6
5
4
3
2
—
27 26 16
MAX_FL
0000_0101_1110_1110
R/W
—
6 5 4 3 2 1 0
FCE BC_REJ PROM MII_MODE DRT LOOP
0000_0000_0000_0001
R/W
IPSBAR + 0x1084
Figure 18-12. Receive Control Register (RCR)
Table 18-21. RCR Field Descriptions
Name
—
MAX_FL
—
FCE
BC_REJ
PROM
MII_MODE
Description
Reserved, should be cleared.
Maximum frame length. Resets to decimal 1518. Length is measured starting at DA and includes the CRC at the end of the frame. Transmit frames longer than MAX_FL will cause the BABT interrupt to occur. Receive Frames longer than
MAX_FL will cause the BABR interrupt to occur and will set the LG bit in the end of frame receive buffer descriptor. The recommended default value to be programmed by the user is 1518 or 1522 (if VLAN Tags are supported).
Reserved, should be cleared.
Flow control enable. If asserted, the receiver will detect
PAUSE frames. Upon PAUSE frame detection, the transmitter will stop transmitting data frames for a given duration.
Broadcast frame reject. If asserted, frames with DA
(destination address) = FF_FF_FF_FF_FF_FF will be rejected unless the PROM bit is set. If both BC_REJ and
PROM = 1, then frames with broadcast DA will be accepted and the M (MISS) bit will be set in the receive buffer descriptor.
Promiscuous mode. All frames are accepted regardless of address matching.
Media independent interface mode. Selects external interface mode. Setting this bit to one selects MII mode, setting this bit equal to zero selects 7-wire mode (used only for serial 10 Mbps). This bit controls the interface mode for both transmit and receive blocks.
18-30
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Bits
1
0
Programming Model
Table 18-21. RCR Field Descriptions (continued)
Name
DRT
LOOP
Description
Disable receive on transmit.
0 Receive path operates independently of transmit (use for full duplex or to monitor transmit activity in half duplex mode).
1 Disable reception of frames while transmitting (normally used for half duplex mode).
Internal loopback. If set, transmitted frames are looped back internal to the device and the transmit output signals are not asserted. The system clock is substituted for the ETXCLK when LOOP is asserted. DRT must be set to zero when asserting LOOP.
18.5.4.10 Transmit Control Register (TCR)
The TCR is read/write and is written by the user to configure the transmit block. This register is cleared at system reset. Bits 2 and 1 should be modified only when ECR[ETHER_EN] = 0.
Field
Reset
R/W
31
Field
Reset
R/W
Address
15
16
—
0000_0000_0000_0000
R/W
—
R/W
5 4 3 2 1 0
RFC_PAUSE TFC_PAUSE FDEN HBC GTS
0000_0000_0000_0000
R
IPSBAR + 0x10C4
R/W
Figure 18-13. Transmit Control Register (TCR)
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-31
Fast Ethernet Controller (FEC)
Bits
31–5
4
3
2
1
0
Table 18-22. TCR Field Descriptions
Name Description
— Reserved, should be cleared.
RFC_PAUSE Receive frame control pause. This read-only status bit will be asserted when a full duplex flow control pause frame has been received and the transmitter is paused for the duration defined in this pause frame. This bit will automatically clear when the pause duration is complete.
TFC_PAUSE Transmit frame control pause. Transmits a PAUSE frame when asserted. When this bit is set, the MAC will stop transmission of data frames after the current transmission is complete. At this time, the GRA interrupt in the EIR register will be asserted. With transmission of data frames stopped, the MAC will transmit a
MAC Control PAUSE frame. Next, the MAC will clear the
TFC_PAUSE bit and resume transmitting data frames. Note that if the transmitter is paused due to user assertion of GTS or reception of a PAUSE frame, the MAC may still transmit a MAC
Control PAUSE frame.
FDEN Full duplex enable. If set, frames are transmitted independent of carrier sense and collision inputs. This bit should only be modified when ETHER_EN is deasserted.
HBC
GTS
Heartbeat control. If set, the heartbeat check is performed following end of transmission and the HB bit in the status register will be set if the collision input does not assert within the heartbeat window. This bit should only be modified when
ETHER_EN is deasserted.
Graceful transmit stop. When this bit is set, the MAC will stop transmission after any frame that is currently being transmitted is complete and the GRA interrupt in the EIR register will be asserted. If frame transmission is not currently underway, the
GRA interrupt will be asserted immediately. Once transmission has completed, a “restart” can be accomplished by clearing the
GTS bit. The next frame in the transmit FIFO will then be transmitted. If an early collision occurs during transmission when
GTS = 1, transmission will stop after the collision. The frame will be transmitted again once GTS is cleared. Note that there may be old frames in the transmit FIFO that will be transmitted when
GTS is reasserted. To avoid this deassert ECR[ETHER_EN] following the GRA interrupt.
18.5.4.11 Physical Address Low Register (PALR)
The PALR is written by the user. This register contains the lower 32 bits (bytes 0,1,2,3) of the 48-bit address used in the address recognition process to compare with the DA (Destination Address) field of receive frames with an individual DA. In addition, this register is used in bytes 0 through 3 of the 6-byte source address field when transmitting PAUSE frames. This register is not reset and must be initialized by the user.
18-32
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Programming Model
Field
Reset
R/W
31
Field
Reset
R/W
Address
15
Bits
31–0
PADDR1
Uninitialized
R/W
PADDR1
Uninitialized
R/W
IPSBAR + 0x10E4
Figure 18-14. Physical Address Low Register (PALR)
Table 18-23. PALR Field Descriptions
Name
PADDR1
Description
Bytes 0 (bits 31:24), 1 (bits 23:16), 2 (bits 15:8) and 3 (bits
7:0) of the 6-byte individual address to be used for exact match, and the Source Address field in PAUSE frames.
16
0
18.5.4.12 Physical Address High Register (PAUR)
The PAUR is written by the user. This register contains the upper 16 bits (bytes 4 and 5) of the 48-bit address used in the address recognition process to compare with the DA (destination address) field of receive frames with an individual DA. In addition, this register is used in bytes 4 and 5 of the 6-byte Source
Address field when transmitting PAUSE frames. Bits 15:0 of PAUR contain a constant type field (0x8808) used for transmission of PAUSE frames. This register is not reset and bits 31:16 must be initialized by the user.
16
Field
Reset
R/W
31
Field
Reset
R/W
Address
15
PADDR2
Uninitialized
R/W
TYPE
1000_1000_0000_1000
R
IPSBAR + 0x10E8
Figure 18-15. Physical Address High Register (PAUR)
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-33
Fast Ethernet Controller (FEC)
BIts
31–16
15–0
Table 18-24. PAUR Field Descriptions
Name
PADDR2
TYPE
Description
Bytes 4 (bits 31:24) and 5 (bits 23:16) of the 6-byte individual address to be used for exact match, and the Source Address field in PAUSE frames.
Type field in PAUSE frames. These 16-bits are a constant value of 0x8808.
18.5.4.13 Opcode/Pause Duration Register (OPD)
The OPD is read/write accessible. This register contains the 16-bit opcode and 16-bit pause duration fields used in transmission of a PAUSE frame. The opcode field is a constant value, 0x0001. When another node detects a PAUSE frame, that node will pause transmission for the duration specified in the pause duration field. This register is not reset and must be initialized by the user.
31 16
Field
Reset
R/W
OPCODE
0000_0000_0000_0001
R
15 0
Field
Reset
R/W
Address
Bits
31–16
15–0
PAUSE_DUR
Uninitialized
R/W
IPSBAR + 0x10EC
Figure 18-16. Opcode/Pause Duration Register (OPD)
Table 18-25. OPD Field Descriptions
Name
OPCODE
PAUSE_DUR
Description
Opcode field used in PAUSE frames.
These bits are a constant, 0x0001.
Pause Duration field used in PAUSE frames.
18.5.4.14 Descriptor Individual Upper Address Register (IAUR)
The IAUR is written by the user. This register contains the upper 32 bits of the 64-bit individual address hash table used in the address recognition process to check for possible match with the DA field of receive frames with an individual DA. This register is not reset and must be initialized by the user.
18-34
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Programming Model
31
Field
Reset
R/W
IADDR1
Uninitialized
R/W
Field
Reset
R/W
Address
15
IADDR1
Uninitialized
R/W
IPSBAR + 0x1118
Figure 18-17. Descriptor Individual Upper Address Register (IAUR)
Bits
31–0
Table 18-26. IAUR Field Descriptions
Name
IADDR1
Descriptions
The upper 32 bits of the 64-bit hash table used in the address recognition process for receive frames with a unicast address. Bit 31 of IADDR1 contains hash index bit
63. Bit 0 of IADDR1 contains hash index bit 32.
16
0
18.5.4.15 Descriptor Individual Lower Address (IALR)
The IALR register is written by the user. This register contains the lower 32 bits of the 64-bit individual address hash table used in the address recognition process to check for possible match with the DA field of receive frames with an individual DA. This register is not reset and must be initialized by the user.
31 16
Field
Reset
R/W
IADDR2
Uninitialized
R/W
15 0
Field
Reset
R/W
Address
IADDR2
Uninitialized
R/W
IPSBAR + 0x111C
Figure 18-18. Descriptor Individual Lower Address Register (IALR)
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-35
Fast Ethernet Controller (FEC)
Bits
31–0
Table 18-27. IALR Field Descriptions
Name
IADDR2
Description
The lower 32 bits of the 64-bit hash table used in the address recognition process for receive frames with a unicast address. Bit 31 of IADDR2 contains hash index bit
31. Bit 0 of IADDR2 contains hash index bit 0.
18.5.4.16 Descriptor Group Upper Address (GAUR)
The GAUR is written by the user. This register contains the upper 32 bits of the 64-bit hash table used in the address recognition process for receive frames with a multicast address. This register must be initialized by the user.
16 31
Field
Reset
R/W
GADDR1
Uninitialized
R/W
Field
Reset
R/W
Address
15
Bits
31–0
GADDR1
Uninitialized
R/W
IPSBAR + 0x1120
Figure 18-19. Descriptor Group Upper Address Register (GAUR)
Table 18-28. GAUR Field Descriptions
Name
GADDR1
Description
The GADDR1 register contains the upper 32 bits of the 64-bit hash table used in the address recognition process for receive frames with a multicast address. Bit 31 of GADDR1 contains hash index bit 63. Bit 0 of GADDR1 contains hash index bit 32.
0
18.5.4.17 Descriptor Group Lower Address (GALR)
The GALR register is written by the user. This register contains the lower 32 bits of the 64-bit hash table used in the address recognition process for receive frames with a multicast address. This register must be initialized by the user.
18-36
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Programming Model
31
Field
Reset
R/W
GADDR2
Uninitialized
R/W
Field
Reset
R/W
Address
15
Bits
31–0
GADDR2
Uninitialized
R/W
IPSBAR + 0x1124
Figure 18-20. Descriptor Group Lower Address Register (GALR)
Table 18-29. GALR Field Descriptions
Name
GADDR2
Description
The GADDR2 register contains the lower 32 bits of the 64-bit hash table used in the address recognition process for receive frames with a multicast address. Bit 31 of GADDR2 contains hash index bit 31. Bit 0 of GADDR2 contains hash index bit 0.
16
0
18.5.4.18 FIFO Transmit FIFO Watermark Register (TFWR)
The TFWR is a 2-bit read/write register programmed by the user to control the amount of data required in the transmit FIFO before transmission of a frame can begin. This allows the user to minimize transmit latency (TFWR = 0x) or allow for larger bus access latency (TFWR = 11) due to contention for the system bus. Setting the watermark to a high value will minimize the risk of transmit FIFO underrun due to contention for the system bus. The byte counts associated with the TFWR field may need to be modified to match a given system requirement (worst case bus access latency by the transmit data DMA channel).
31 16
Field
Reset
R/W
—
0000_0000_0000_0000
R/W
15 2 1 0
X_WMRK Field
Reset
R/W
Address
—
0000_0000_0000_0000
R/W
IPSBAR + 0x1144
Figure 18-21. FIFO Transmit FIFO Watermark Register (TFWR)
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-37
Fast Ethernet Controller (FEC)
Bits
31–2
1–0
Table 18-30. TFWR Field Descriptions
Name
—
X_WMRK
Descriptions
Reserved, should be cleared.
Number of bytes written to transmit FIFO before transmission of a frame begins
0x 64 bytes written
10 128 bytes written
11 192 bytes written
18.5.4.19 FIFO Receive Bound Register (FRBR)
The FRBR is an 8-bit register that the user can read to determine the upper address bound of the FIFO
RAM. Drivers can use this value, along with the FRSR to appropriately divide the available FIFO RAM between the transmit and receive data paths.
31 16
Field
Reset
R/W
Field
Reset
R/W
Address
15
Bits
31–10
9–2
1–0
—
0000_0000_0000_0000
Read Only
—
10 9
R_BOUND
0000_0110_0000_0000
Read Only
IPSBAR + 0x114C
Figure 18-22. FIFO Receive Bound Register (FRBR)
Table 18-31. FRBR Field Descriptions
Name
—
R_BOUND
—
2
Descriptions
Reserved, read as 0 (except bit 10, which is read as 1).
Read-only. Highest valid FIFO RAM address.
Reserved, should be cleared.
1
—
0
18-38
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Programming Model
18.5.4.20 FIFO Receive Start Register (FRSR)
The FRSR is an 8-bit register programmed by the user to indicate the starting address of the receive FIFO.
FRSR marks the boundary between the transmit and receive FIFOs. The transmit FIFO uses addresses from the start of the FIFO to the location four bytes before the address programmed into the FRSR. The receive FIFO uses addresses from FRSR to FRBR inclusive.
The FRSR register is initialized by hardware at reset. FRSR only needs to be written to change the default value.
31 16
Field
Reset
R/W
Field
Reset
R/W
Address
15
Bits
31–10
9–2
1–0
—
0000_0000_0000_0000
R/W
—
10 9
R_FSTART
0000_0101_0000_0000
R/W
IPSBAR + 0x1150
Figure 18-23. FIFO Receive Start Register (FRSR)
Table 18-32. FRSR Field Descriptions
Name
—
R_FSTART
—
2
Descriptions
Reserved, read as 0 (except bit 10, which is read as 1).
Address of first receive FIFO location. Acts as delimiter between receive and transmit FIFOs.
Reserved, read as 0.
1
—
0
18.5.4.21 Receive Descriptor Ring Start (ERDSR)
The ERDSR is written by the user. It provides a pointer to the start of the circular receive buffer descriptor queue in external memory. This pointer must be 32-bit aligned; however, it is recommended it be made
128-bit aligned (evenly divisible by 16).
This register is not reset and must be initialized by the user prior to operation.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-39
Fast Ethernet Controller (FEC)
31
Field
Reset
R/W
R_DES_START
Uninitialized
R/W
Field
Reset
R/W
Address
15
Bits
31–2
1–0
R_DES_START
Uninitialized
R/W
IPSBAR + 0x1180
Figure 18-24. Receive Descriptor Ring Start Register (ERDSR)
Table 18-33. ERDSR Field Descriptions
Name Descriptions
R_DES_START Pointer to start of receive buffer descriptor queue.
— Reserved, should be cleared.
2
16
1
—
0
18.5.4.22 Transmit Buffer Descriptor Ring Start (ETSDR)
The ETSDR is written by the user. It provides a pointer to the start of the circular transmit buffer descriptor queue in external memory. This pointer must be 32-bit aligned; however, it is recommended it be made
128-bit aligned (evenly divisible by 16). Bits 1 and 0 should be written to 0 by the user. Non-zero values in these two bit positions are ignored by the hardware.
This register is not reset and must be initialized by the user prior to operation.
16 31
Field
Reset
R/W
X_DES_START
Uninitialized
R/W
Field
Reset
R/W
Address
15
X_DES_START
Uninitialized
R/W
IPSBAR + 0x1184
Figure 18-25. Transmit Buffer Descriptor Ring Start Register (ETDSR)
2 1
—
0
18-40
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Programming Model
Bits
31–2
1–0
Table 18-34. ETDSR Field Descriptions
Name Descriptions
X_DES_START Pointer to start of transmit buffer descriptor queue.
— Reserved, should be cleared.
18.5.4.23 Receive Buffer Size Register (EMRBR)
The EMRBR is a 9-bit register programmed by the user. The EMRBR register dictates the maximum size of all receive buffers. Note that because receive frames will be truncated at 2k-1 bytes, only bits 10–4 are used. This value should take into consideration that the receive CRC is always written into the last receive buffer. To allow one maximum size frame per buffer, EMRBR must be set to RCR[MAX_FL] or larger.
The EMRBR must be evenly divisible by 16. To insure this, bits 3-0 are forced low. To minimize bus utilization (descriptor fetches) it is recommended that EMRBR be greater than or equal to 256 bytes.
The EMRBR register does not reset, and must be initialized by the user.
31 16
Field
Reset
R/W
Field
Reset
R/W
Address
15
Bits
30–11
10–4
3–0
—
Uninitialized
R/W
11 10 4
— R_BUF_SIZE
Uninitialized
R/W
IPSBAR + 0x11B8
Figure 18-26. Receive Buffer Size Register (EMRBR)
3
Table 18-35. EMRBR Field Descriptions
Name Descriptions
— Reserved, should be written to 0 by the host processor.
R_BUF_SIZE Receive buffer size.
— Reserved, should be written to 0 by the host processor.
—
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-41
Fast Ethernet Controller (FEC)
18.6
Buffer Descriptors
This section provides a description of the operation of the driver/DMA via the buffer descriptors. It is followed by a detailed description of the receive and transmit descriptor fields.
18.6.1
Driver/DMA Operation with Buffer Descriptors
The data for the FEC frames must reside in memory external to the FEC. The data for a frame is placed in one or more buffers. Associated with each buffer is a buffer descriptor (BD) which contains a starting address (pointer), data length, and status/control information (which contains the current state for the buffer). To permit maximum user flexibility, the BDs are also located in external memory and are read in by the FEC DMA engine.
Software “produces” buffers by allocating/initializing memory and initializing buffer descriptors. Setting the RxBD[E] or TxBD[R] bit “produces” the buffer. Software writing to either the TDAR or RDAR tells the FEC that a buffer has been placed in external memory for the transmit or receive data traffic, respectively. The hardware reads the BDs and “consumes” the buffers after they have been produced. After the data DMA is complete and the buffer descriptor status bits have been written by the DMA engine, the
RxBD[E] or TxBD[R] bit will be cleared by hardware to signal the buffer has been “consumed.” Software may poll the BDs to detect when the buffers have been consumed or may rely on the buffer/frame interrupts. These buffers may then be processed by the driver and returned to the free list.
The ECR[ETHER_EN] signal operates as a reset to the BD/DMA logic. When ECR[ETHER_EN] is deasserted the DMA engine BD pointers are reset to point to the starting transmit and receive BDs. The buffer descriptors are not initialized by hardware during reset. At least one transmit and receive buffer descriptor must be initialized by software before the ECR[ETHER_EN] bit is set.
The buffer descriptors operate as two separate rings. ERDSR defines the starting address for receive BDs and ETDSR defines the starting address for transmit BDs. The last buffer descriptor in each ring is defined by the Wrap (W) bit. When set, W indicates that the next descriptor in the ring is at the location pointed to by ERDSR and ETDSR for the receive and transmit rings, respectively. Buffer descriptor rings must start on a 32-bit boundary; however, it is recommended they are made 128-bit aligned.
18.6.1.1
Driver/DMA Operation with Transmit BDs
Typically a transmit frame will be divided between multiple buffers. An example is to have an application payload in one buffer, TCP header in a 2nd buffer, IP header in a 3rd buffer, Ethernet/IEEE 802.3 header in a 4th buffer. The Ethernet MAC does not prepend the Ethernet header (destination address, source address, length/type field(s)), so this must be provided by the driver in one of the transmit buffers. The
Ethernet MAC can append the Ethernet CRC to the frame. Whether the CRC is appended by the MAC or by the driver is determined by the TC bit in the transmit BD which must be set by the driver.
The driver (TxBD software producer) should set up Tx BDs in such a way that a complete transmit frame is given to the hardware at once. If a transmit frame consists of three buffers, the BDs should be initialized with pointer, length and control (W, L, TC, ABC) and then the TxBD[R] bits should be set = 1 in reverse order (3rd, 2nd, 1st BD) to insure that the complete frame is ready in memory before the DMA begins. If the TxBDs are set up in order, the DMA Controller could DMA the first BD before the 2nd was made available, potentially causing a transmit FIFO underrun.
In the FEC, the DMA is notified by the driver that new transmit frame(s) are available by writing to the
TDAR register. When this register is written to (data value is not significant) the FEC RISC will tell the
DMA to read the next transmit BD in the ring. Once started, the RISC + DMA will continue to read and interpret transmit BDs in order and DMA the associated buffers, until a transmit BD is encountered with the R bit = 0. At this point the FEC will poll this BD one more time. If the R bit = 0 the second time, then
18-42
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Buffer Descriptors the RISC will stop the transmit descriptor read process until software sets up another transmit frame and writes to TDAR.
When the DMA of each transmit buffer is complete, the DMA writes back to the BD to clear the R bit, indicating that the hardware consumer is finished with the buffer.
18.6.1.2
Driver/DMA Operation with Receive BDs
Unlike transmit, the length of the receive frame is unknown by the driver ahead of time. Therefore the driver must set a variable to define the length of all receive buffers. In the FEC, this variable is written to the EMRBR register.
The driver (RxBD software producer) should set up some number of “empty” buffers for the Ethernet by initializing the address field and the E and W bits of the associated receive BDs. The hardware (receive
DMA) will consume these buffers by filling them with data as frames are received and clearing the E bit and writing to the L (1 indicates last buffer in frame) bit, the frame status bits (if L = 1) and the length field.
If a receive frame spans multiple receive buffers, the L bit is only set for the last buffer in the frame. For non-last buffers, the length field in the receive BD will be written by the DMA (at the same time the E bit is cleared) with the default receive buffer length value. For end of frame buffers the receive BD will be written with L = 1 and information written to the status bits (M, BC, MC, LG, NO, CR, OV, TR). Some of the status bits are error indicators which, if set, indicate the receive frame should be discarded and not given to higher layers. The frame status/length information is written into the receive FIFO following the end of the frame (as a single 32-bit word) by the receive logic. The length field for the end of frame buffer will be written with the length of the entire frame, not just the length of the last buffer.
For simplicity the driver may assign the default receive buffer length to be large enough to contain an entire frame, keeping in mind that a malfunction on the network or out of spec implementation could result in giant frames. Frames of 2k (2048) bytes or larger are truncated by the FEC at 2047 bytes so software is guaranteed never to see a receive frame larger than 2047 bytes.
Similar to transmit, the FEC will poll the receive descriptor ring after the driver sets up receive BDs and writes to the RDAR register. As frames are received the FEC will fill receive buffers and update the associated BDs, then read the next BD in the receive descriptor ring. If the FEC reads a receive BD and finds the E bit = 0, it will poll this BD once more. If the BD = 0 a second time the FEC will stop reading receive BDs until the driver writes to RDAR.
18.6.2
Ethernet Receive Buffer Descriptor (RxBD)
In the RxBD, the user initializes the E and W bits in the first longword and the pointer in second longword.
When the buffer has been DMA’d, the Ethernet controller will modify the E, L, M, BC, MC, LG, NO, CR,
OV, and TR bits and write the length of the used portion of the buffer in the first longword. The M, BC,
MC, LG, NO, CR, OV and TR bits in the first longword of the buffer descriptor are only modified by the
Ethernet controller when the L bit is set.
Figure 18-27. Receive Buffer Descriptor (RxBD)
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-43
Fast Ethernet Controller (FEC)
Offset + 0
15 14 13 12 11
E RO1 W RO2 L
10
—
Offset + 2
Offset + 4
Offset + 6
9
—
8
M
7 6 5 4 3 2 1 0
BC MC LG NO — CR OV TR
Data Length
Rx Data Buffer Pointer - A[31:16]
Rx Data Buffer Pointer - A[15:0]
Word
Offset + 0
Offset + 0
Offset + 0
Offset + 0
Offset + 0
Offset + 0
Offset + 0
Offset + 0
Offset + 0
Table 18-36. Receive Buffer Descriptor Field Definitions
Location
Bit 15
Bit 14
Bit 13
Bit 12
Bit 11
Bits 10–9
Bit 8
Bit 7
Bit 6
Field Name
E
RO1
W
RO2
L
—
M
BC
MC
Description
Empty. Written by the FEC (=0) and user (=1).
0 The data buffer associated with this BD has been filled with received data, or data reception has been aborted due to an error condition. The status and length fields have been updated as required.
1 The data buffer associated with this BD is empty, or reception is currently in progress.
Receive software ownership.
This field is reserved for use by software. This read/write bit will not be modified by hardware, nor will its value affect hardware.
Wrap. Written by user.
0 The next buffer descriptor is found in the consecutive location
1 The next buffer descriptor is found at the location defined in
ERDSR.
Receive software ownership.
This field is reserved for use by software. This read/write bit will not be modified by hardware, nor will its value affect hardware.
Last in frame. Written by the FEC.
0 The buffer is not the last in a frame.
1 The buffer is the last in a frame.
Reserved.
Miss. Written by the FEC. This bit is set by the FEC for frames that were accepted in promiscuous mode, but were flagged as a “miss” by the internal address recognition. Thus, while in promiscuous mode, the user can use the M-bit to quickly determine whether the frame was destined to this station. This bit is valid only if the L-bit is set and the PROM bit is set.
0 The frame was received because of an address recognition hit.
1 The frame was received because of promiscuous mode.
Will be set if the DA is broadcast (FF-FF-FF-FF-FF-FF).
Will be set if the DA is multicast and not BC.
18-44
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Word
Offset + 0
Offset + 0
Offset + 0
Offset + 0
Offset + 0
Offset + 0
Offset + 2
0ffset + 4
Buffer Descriptors
Table 18-36. Receive Buffer Descriptor Field Definitions (continued)
Location
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Bits [15:0]
Bits [15:0]
Field Name Description
LG
NO
Rx frame length violation. Written by the FEC. A frame length greater than RCR[MAX_FL] was recognized. This bit is valid only if the L-bit is set. The receive data is not altered in any way unless the length exceeds 2047 bytes.
Receive non-octet aligned frame. Written by the FEC. A frame that contained a number of bits not divisible by 8 was received, and the CRC check that occurred at the preceding byte boundary generated an error. This bit is valid only if the L-bit is set. If this bit is set the CR bit will not be set.
—
CR
OV
TR
Reserved.
Receive CRC error. Written by the FEC. This frame contains a
CRC error and is an integral number of octets in length. This bit is valid only if the L-bit is set.
Overrun. Written by the FEC. A receive FIFO overrun occurred during frame reception. If this bit is set, the other status bits, M,
LG, NO, CR, and CL lose their normal meaning and will be zero. This bit is valid only if the L-bit is set.
Will be set if the receive frame is truncated (frame length >
2047 bytes). If the TR bit is set the frame should be discarded and the other error bits should be ignored as they may be incorrect.
Data Length Data length. Written by the FEC. Data length is the number of octets written by the FEC into this BD’s data buffer if L = 0 (the value will be equal to EMRBR), or the length of the frame including CRC if L = 1. It is written by the FEC once as the BD is closed.
A[31:16] RX data buffer pointer, bits [31:16]
1
Offset + 6 Bits [15:0] A[15:0] RX data buffer pointer, bits [15:0]
1
The receive buffer pointer, which contains the address of the associated data buffer, must always be evenly divisible by 16. The buffer must reside in memory external to the FEC. This value is never modified by the
Ethernet controller.
NOTE
Whenever the software driver sets an E bit in one or more receive descriptors, the driver should follow that with a write to RDAR.
18.6.3
Ethernet Transmit Buffer Descriptor (TxBD)
Data is presented to the FEC for transmission by arranging it in buffers referenced by the channel’s TxBDs.
The Ethernet controller confirms transmission by clearing the ready bit (R bit) when DMA of the buffer is complete. In the TxBD the user initializes the R, W, L, and TC bits and the length (in bytes) in the first longword, and the buffer pointer in the second longword.
The FEC will set the R bit = 0 in the first longword of the BD when the buffer has been DMA’d. Status bits for the buffer/frame are not included in the transmit buffer descriptors. Transmit frame status is
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-45
Fast Ethernet Controller (FEC) indicated via individual interrupt bits (error conditions) and in statistic counters in the MIB block. See
Section 18.5.3, “MIB Block Counters Memory Map
” for more details.
3 2 1 0
Offset + 0
15 14 13 12
R TO1 W TO2
11
L
Offset + 2
10 9
TC ABC
8 7
Data Length
6
Offset + 4
5
Tx Data Buffer Pointer - A[31:16]
Offset + 6
Tx Data Buffer Pointer - A[15:0]
4
Word
Offset + 0
Offset + 0
Offset + 0
Offset + 0
Offset + 0
Offset + 0
Offset + 0
Offset + 0
Figure 18-28. Transmit Buffer Descriptor (TxBD)
Table 18-37. Transmit Buffer Descriptor Field Definitions
Location
Bit 15
Bit 14
Bit 13
BIt 12
Bit 11
Bit 10
Bit 9
Bits [8:0]
Field Name
R
TO1
W
TO2
L
TC
ABC
—
Description
Ready. Written by the FEC and the user.
0 The data buffer associated with this BD is not ready for transmission. The user is free to manipulate this BD or its associated data buffer. The FEC clears this bit after the buffer has been transmitted or after an error condition is encountered.
1 The data buffer, which has been prepared for transmission by the user, has not been transmitted or is currently being transmitted. No fields of this BD may be written by the user once this bit is set.
Transmit software ownership. This field is reserved for software use. This read/write bit will not be modified by hardware, nor will its value affect hardware.
Wrap. Written by user.
0 The next buffer descriptor is found in the consecutive location
1 The next buffer descriptor is found at the location defined in
ETDSR.
Transmit software ownership. This field is reserved for use by software. This read/write bit will not be modified by hardware, nor will its value affect hardware.
Last in frame. Written by user.
0 The buffer is not the last in the transmit frame.
1 The buffer is the last in the transmit frame.
Tx CRC. Written by user (only valid if L = 1).
0 End transmission immediately after the last data byte.
1 Transmit the CRC sequence after the last data byte.
Append bad CRC. Written by user (only valid if L = 1).
0 No effect
1 Transmit the CRC sequence inverted after the last data byte
(regardless of TC value).
Reserved.
18-46
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Buffer Descriptors
Table 18-37. Transmit Buffer Descriptor Field Definitions (continued)
Word Location Field Name Description
Offset + 2
Offset + 4
Bits [15:0]
Bits [15:0]
Data Length Data Length, written by user.
Data length is the number of octets the FEC should transmit from this BD’s data buffer. It is never modified by the FEC. Bits
[15:5] are used by the DMA engine, bits[4:0] are ignored.
A[31:16] Tx data buffer pointer, bits [31:16]
1
Offset + 6 Bits [15:0] A[15:0] Tx data buffer pointer, bits [15:0].
1
The transmit buffer pointer, which contains the address of the associated data buffer, must always be evenly divisible by 4. The buffer must reside in memory external to the FEC. This value is never modified by the
Ethernet controller.
NOTE
Once the software driver has set up the buffers for a frame, it should set up the corresponding BDs. The last step in setting up the BDs for a transmit frame should be to set the R bit in the first BD for the frame. The driver should follow that with a write to TDAR which will trigger the FEC to poll the next BD in the ring.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 18-47
Fast Ethernet Controller (FEC)
18-48
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Chapter 19
Ethernet Physical Transceiver (EPHY) Block Description
19.1
Introduction
The Ethernet physical transceiver (Ethernet physical interface) is an IEEE 802.3 compliant
10BASE-T/100BASE-TX Ethernet PHY transceiver. The Ethernet physical interface module supports both the medium-independent interface (MII) and the MII management interface. The EPHY requires a
25-MHz crystal for its basic operation.
19.1.1
Features
• IEEE 802.3 compliant
• Full-/half-duplex support in all modes
• Medium-independent interface (MII), which has these characteristics:
— Capable of supporting both 10 Mbps and 100 Mbps data rates
— Data and delimiters are synchronous to clock references
— Provides independent four-bit wide transmit and receive data paths
— Provides a simple management interface
• Supports auto-negotiation
• Auto-negotiation next page ability
• Single RJ45 connection
• 1:1 common transformer
• Baseline wander correction
• Digital adaptive equalization
• Integrated wave-shaping circuitry
• Far-end fault detect
• MDC rates up to 25 MHz
• Supports MDIO preamble suppression
• Jumbo packet
• 2.5 V CMOS
• 2.5 V MII interface
• 125 MHz clock generator and timing recovery
• Loopback modes
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
19-1
Ethernet Physical Transceiver (EPHY) Block Description
.
19.1.2
Block Diagram
PHY_TXP
PHY_TXN
PHY_RXP
PHY_RXN
PHY_RBIAS
PHY SUB BLOCK
IP BUS
REGISTERS
Figure 19-1. Ethernet Physical Transceiver (EPHY) Block Diagram
MII_RXCLK
MII_RXDV
MII_RXD[3:0]
MII_RXER
MII_TXCLK
MII_TXEN
MII_TXD[3:0]
MII_TXER
MII_CRS
MII_COL
MII_MDC
MII_MDIO
MII INTERFACE
IP BUS
SIGNALS
REF CLOCK
19-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
External Signal Descriptions
RXP
RXN
10BASE-T
RECEIVER
POLARITY CORRECTION
SQUELCH
LINK DETECT
CLOCK RECOVERY
MANCHESTER DECODE
TXP
100BASE-TX
LOOPBACK
TXN
RBIAS
100BASE-TX
RECEIVER
10BASE-T
DRIVER
100BASE-TX
DRIVER
VGA CONTROL
(COARSE EQUALIZER)
DIGITAL EQUALIZER
SLICER
TIMING CONTROL
BLW CONTROL
10BASE-T
DIG LOOP B
MLT-3 DECODE
DESCRAMBLER
AUTO
NEGOTIATE
4B/5B
DECODE
COLLISION
CARRIER SENSE
MANCHESTER ENCODER
DIGITAL WAVE SHAPING
100BASE-TX
DIG LOOP B
SCRAMBLER
MLT-3 ENCODE
4B / 5B
ENCODE
MII
LOOPBACK
MII
VOLTAGE/ CURRENT
REFERENCES
10BASE-T
PLL
100BASE-TX
PLL
REF
CLOCK
MANAGEMENT
(MII)
CONFIGURATION
REGISTERS
MDIO
Figure 19-2. PHY Sub Block Diagram
19.2
External Signal Descriptions
This section contains the EPHY external pin descriptions.
19.2.1
PHY_TXP — EPHY Twisted Pair Output +
Ethernet twisted-pair output pin. This pin is high-impedance out of reset.
19.2.2
PHY_TXN — EPHY Twisted Pair Output –
Ethernet twisted-pair output pin. This pin is high-impedance out of reset.
19.2.3
PHY_RXP — EPHY Twisted Pair Input +
Ethernet twisted-pair input pin. This pin is high-impedance out of reset.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
19-3
Ethernet Physical Transceiver (EPHY) Block Description
19.2.4
PHY_RXN — EPHY Twisted Pair Input –
Ethernet twisted-pair input pin. This pin is high-impedance out of reset.
19.2.5
PHY_RBIAS — EPHY Bias Control Resistor
Connect a 1.0% external resistor, RBIAS (see Electrical Characteristics chapter), between the
PHY_RBIAS pin and analog ground. Place this resistor as near to the chip pin as possible. Stray capacitance must be kept to less than 10 pF (>50 pF will cause instability). No high-speed signals are permitted in the region of RBIAS.
19.2.6
PHY_VDDRX, PHY_VSSRX — Power Supply Pins for EPHY Receiver
Power is supplied to the EPHY receiver through PHY_VDDRX and PHY_VSSRX. This 2.5 V supply is derived from the internal voltage regulator. There is no static load on those pins allowed. The internal voltage regulator is turned off, if V
DDR
is tied to ground.
19.2.7
PHY_VDDTX, PHY_VSSTX — Power Supply Pins for EPHY
Transmitter
External power is supplied to the EPHY transmitter through PHY_VDDTX and PHY_VSSTX. This 2.5 V supply is derived from the internal voltage regulator. There is no static load on those pins allowed. The internal voltage regulator is turned off, if V
DDR
is tied to ground.
19.2.8
PHY_VDDA, PHY_VSSA — Power Supply Pins for EPHY Analog
Power is supplied to the EPHY PLLs through PHY_VDDA and PHY_VSSA. This 2.5 V supply is derived from the internal voltage regulator. There is no static load on those pins allowed. The internal voltage regulator is turned off, if V
DDR
is tied to ground.
19.2.9
COLLED — Collision LED
Flashes in half-duplex mode when a collision occurs on the network if EPHYCTL0 LEDEN bit is set.
19.2.10 DUPLED — Duplex LED
Indicates the duplex of the link, which can be full-duplex or half-duplex if EPHYCTL0 LEDEN bit is set.
19.2.11 SPDLED — Speed LED
Indicates the speed of a link, which can be 10 Mbps or 100 Mbps if EPHYCTL0 LEDEN bit is set.
19.2.12 LNKLED — Link LED
Indicates whether a link is established with another network device if EPHYCTL0 LEDEN bit is set.
19-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Memory Map and Register Descriptions
19.2.13 ACTLEC — Activity LED
Flashes when data is received by the device if EPHYCTL0 LEDEN bit is set.
19.3
Memory Map and Register Descriptions
This section provides a detailed description of all registers accessible in the Ethernet physical interface.
19.3.1
Module Memory Map
gives an overview of all registers in the Ethernet physical interface memory map. The Ethernet physical interface occupies 48 bytes in the memory space. The register address results from the addition of base address and address offset.
The base address is determined at the MCU level. The address offset is defined at the module level.
Table 19-1. EPHY Module Memory Map
Address
Offset
$__00
$__01
$__02
$__03
Use
Ethernet Physical Transceiver Control Register 0 (EPHYCTL0)
Ethernet Physical Transceiver Control Register 1 (EPHYCTL1)
Ethernet Physical Transceiver Status Register (EPHYSR)
RESERVED
Access
R/W
R/W
R/W
R
19.3.2
Register Descriptions
19.3.2.1
Ethernet Physical Transceiver Control Register 0 (EPHYCTL0)
Module Base + $0x0000
7 6
R
W
RESET:
EPHYEN
0
ANDIS
1
5
DIS100
1
4
DIS10
1
3 2
LEDEN EPHYWAI
0 0
1
0
0
0
EPHYIEN
0
= Unimplemented or Reserved
Figure 19-3. Ethernet Physical Transceiver Control Register 0 (EPHYCTL0)
Read: Anytime
Write: See each bit description
EPHYEN — EPHY Enable
This bit can be written anytime.
1 = Enables EPHY
0 = Disables EPHY
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
19-5
Ethernet Physical Transceiver (EPHY) Block Description
ANDIS — Auto Negotiation Disable
This bit can be written anytime, but the value is latched in the ANE bit of the MII PHY control register
(MII address 0.12) only when the EPHYEN bit transitions from 0 to 1.
1 = Auto negotiation is disabled after start-up. A 0 is latched in the ANE bit of the MII PHY control register (MII address 0.12), and upon completion of the start-up delay (t
Start-up
), the EPHY will bypass auto-negotiation. The mode of operation will be determined by the manual setting of
MII registers.
0 = Auto negotiation is enabled after start-up. A 1 is latched in the ANE bit of the MII PHY control register (MII address 0.12), and upon completion of the start-up delay (t
Start-up
), the EPHY will enter auto-negotiation. The mode of operation will be automatically determined.
DIS100 — Disable 100 BASE-TX PLL
This bit can be written anytime. Allows user to power down the clock generation PLL for
100BASE-TX clocks.
1 = Disables 100BASE-TX PLL
0 = 100BASE-TX PLL state determined by EPHY operation mode
DIS10 — Disable 10BASE-T PLL
This bit can be written anytime. Allows user to power down the clock generation PLL for 10BASE-T clocks.
1 = Disables 10BASE-T PLL
0 = 10 BASE-T PLL state determined by EPHY operation mode
LEDEN — LED Drive Enable
This bit can be written anytime.
1 = Enables the EPHY to drive LED signals.
0 = Disables the EPHY to drive LED signals.
EPHYWAI — EPHY Module Stops While in Wait
This bit can be written anytime.
1 = Disables the EPHY module while the MCU is in wait mode. EPHY interrupts cannot be used to bring the MCU out of wait.
0 = Allows the EPHY module to continue running during wait.
EPHYIEN — EPHY Interrupt Enable
This bit can be written anytime.
1 = Enables EPHY module interrupts
0 = Disables EPHY module interrupts
19.3.2.2
Ethernet Physical Transceiver Control Register 1 (EPHYCTL1)
Module Base + $0x0001
7
0
R
W
RESET: 0
6
0
0
5
0
0
4 3 2 1 0
PHYADD4 PHYADD3 PHYADD2 PHYADD1 PHYADD0
0 0 0 0 0
= Unimplemented or Reserved
Figure 19-4. Ethernet Physical Transceiver Control Register 1 (EPHYCTL1)
19-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Memory Map and Register Descriptions
Read: Anytime
Write: See each bit description
PHYADD[4:0] — EPHY Address for MII Requests
These bits can be written anytime, but the EPHY address is latched to the MII PHY address register
(MII address 21.4:0) only when the EPHYEN bit transitions from 0 to 1. PHYADD4 is the MSB of the of the EPHY address.
19.3.2.3
Ethernet Physical Transceiver Status Register (EPHYSR)
Module Base + $0x0002
7
0
R
W
RESET: 0
6
0
0
5
100DIS
1
4
10DIS
1
3
0
0
2
0
0
1
0
0
0
EPHYIF
0
= Unimplemented or Reserved
Figure 19-5. Ethernet Physical Transceiver Status Register (EPHYSR)
Read: Anytime
Write: See bit descriptions
100DIS — EPHY Port 100BASE-TX mode status
This bit is not writable — read only. Output to indicate EPHY port Base100-TX mode status.
1 = EPHY port 100BASE-TX disabled
0 = EPHY port 100BASE-TX enabled
10DIS — EPHY Port 10BASE-T mode status
This bit is not writable. Output to indicate EPHY port 10BASE-T mode status.
1 = EPHY port 10BASE-T disabled
0 = EPHY port 10BASE-T enabled
EPHYIF — EPHY Interrupt Flag
EPHYIF indicates that interrupt conditions have occurred. To clear the interrupt flag, write a 1 to this bit after reading the interrupt control register via the MII management interface.
1 = EPHY interrupt has occurred
0 = EPHY interrupt has not occurred
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
19-7
Ethernet Physical Transceiver (EPHY) Block Description
12
13
14
15
8
9
10
11
16
17
18
6
7
4
5
2
3
0
1
19.3.3
MII Registers
gives an overview of all registers in the Ethernet physical interface that are accessible via the
MII management interface. These registers are not part of the MCU memory map.
Table 19-2. MII Registers
Address
$0x0000
$0x0001
$0x0002
$0x0003
$0x0004
$0x0005
$0x0006
$0x0007
$0x0008
$0x0009
$0x000A
$0x000B
$0x000C
$0x000D
$0x000E
$0x000F
$0x0010
$0x0011
$0x0012
Use
Control Register
Status Register
PHY Identification Register 1
PHY Identification Register 2
Auto-Negotiation Advertisement Register
Auto-Negotiation Link Partner Ability Register
Auto-Negotiation Expansion Register
Auto-Negotiation Next Page Transmit
RESERVED
RESERVED
RESERVED
RESERVED
RESERVED
RESERVED
RESERVED
RESERVED
Interrupt Control Register
Proprietary Status Register
Proprietary Control Register
Access
Read/Write
Read/Write
4
Read/Write
4
Read/Write
4
Read/Write
Read/Write
4
Read/Write
4
Read/Write
Read/Write
1
Read/Write
1
Read/Write
1
Read/Write
1
Read/Write
1
Read/Write
1
Read/Write
1
Read/Write
1
Read/Write
Read/Write
4
Read/Write
1. Always read $00
2. Writable only in special modes (test_mode = 1)
4. Write has no effect.
NOTE
Bit notation for MII registers is: Bit 20.15 refers to MII register address 20 and bit number 15.
19.3.3.1
EPHY Control Register
MII Register Address 0 ($00000)
15 14 13 12
R
W
RESET:
11 10
RESET
0
LOOP
BACK
DATA
RATE
ANE PDWN ISOL
0 1 X 0
= Unimplemented or Reserved
0
9
RAN
0
8
DPLX
1
7
COL
TEST
0
6
0
0
Figure 19-6. Control Register
Read: Anytime
Write: Anytime
5
0
0
4
0
0
3
0
0
2
0
0
1
0
0
0
0
0
19-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Memory Map and Register Descriptions
RESET — EPHY Reset
Resetting a port is accomplished by setting this bit to 1.
1 = The PHY will reset the port’s status and registers to the default values. The PHY will also reset the PHY to its initial state. After the reset is complete, the PHY clears this bit automatically.
The reset process will be completed within 1.3 ms of this bit being set. While the preamble is suppressed, the management interface must not receive an ST within three MDC clock cycles following a software reset.
0 = No effect
LOOPBACK — Digital Loopback Mode
Determines Digital Loopback Mode
1 = Enables digital loopback mode. Port will be placed in loopback mode. Loopback mode will allow the TXD data to be sent to the RXD data circuitry within 512 bit times. The PHY will be isolated from the medium (no transmit or receive to the medium allowed) and the MII_COL signal will remain de-asserted, unless this bit is set.
0 = Disables digital loopback mode
DATARATE — Speed Selection
The link speed will be selected either through the auto-negotiation process or by manual speed selection. ANE allows manual speed selection while it is set to 0. While auto-negotiation is enabled,
DATARATE can be read or written but its value is not required to reflect speed of the link.
1 = While auto-negotiation is disabled, selects 100 Mbps operation
0 = While auto-negotiation is disabled, selects 10 Mbps operation
ANE — Auto-Negotiation Enable
The ANE bit determines whether the A/N process is enabled. When auto-negotiation is disabled,
DATARATE and DPLX determine the link configuration. While auto-negotiation is enabled, bits
DATARATE and DPLX do not affect the link.
1 = Enables auto-negotiation
0 = Disables auto-negotiation
PDWN — Power Down
When this bit is set, the port is placed in a low power consumption mode.
1 = Port is placed in a low power consumption mode. Normal operation will be allowed within 0.5 s after PDWN and ISOL are changed to 0. During a transition to power-down mode (or if already in power down mode), the port will respond only to management function requests through the
MI interface. All other port operations will be disabled. When power-down mode is exited, all register values are maintained. The port will start its operation based on the register values.
0 = Normal operation
ISOL — Isolate
1 = Isolates the port’s data path signals from the MII. The port will not respond to changes on
MII_TXDx, MII_TXEN, and MII_TXER inputs, and it will present high impedance on
MII_TXCLK, MII_RXCLK, MII_RXDV, MII_RXER, MII_RXDx, MII_COL, and MII_CRS outputs. The port will respond to management transactions while in isolate mode.
0 = Normal operation
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
19-9
Ethernet Physical Transceiver (EPHY) Block Description
RAN — Restart Auto-Negotiation
The RAN bit determines when the A/N process can start processing.
1 = When auto-negotiation is enabled (ANE=1), the auto-negotiation process will be restarted.
After auto-negotiation indicates that it has been initialized, this bit is cleared. When bit ANE is cleared to indicate auto-negotiation is disabled, RAN must also be 0.
0 = Normal operation.
DPLX — Duplex Mode
This mode can be selected by either the auto-negotiation process or manual duplex selection. Manual duplex selection is allowed only while the auto-negotiation process is disabled (ANE=0). While the auto-negotiation process is enabled (ANE = 1), the state of DPLX has no effect on the link configuration. While loopback mode is asserted (LOOPBACK =1), the value of DPLX will have no effect on the PHY.
1 = Indicates full-duplex mode
0 = Indicates half-duplex mode
COLTEST — Collision Test
The collision test function will be enabled only if the loopback mode of operation is also selected
(LOOPBACK = 1).
1 = Forces the PHY to assert the MII_COL signal within 512 bit times from the assertion of
MII_TXEN and de-assert MII_COL within 4 bit times of MII_TXEN being de-asserted.
0 = Normal operation
19.3.3.2
Status Register
This register advertises the capabilities of the port to the MII.
MII Register Address 1 (%00001)
15
R
100
T4
W
RESET: 0
14
100X
FD
1
13
100X
HD
1
12
10T
FD
1
11
10T
HD
1
10
0
0
9
0
0
8
0
0
7
0
0
6
SUP
PRE
5
AN
COMP
4
REM
FLT
3
AN
ABL
2
LNK
STST
1
JAB
DT
1 0 0 1 0 0
0
EX
CAP
1
= Unimplemented or Reserved
Figure 19-7. Status Register
Read: Anytime
Write: Writes have no effect
100T4 —100BASE-T4
1 = Indicates PHY supports 100BASE-T4 transmission
0 = Indicates the PHY does not support 100BASE-T4 transmission
This function is not implemented in the EPHY module.
100XFD —100BASE-TX Full-Duplex
1 = Indicates PHY supports 100BASE-TX full-duplex mode
0 = Indicates PHY does not support 100BASE-TX full-duplex mode
19-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Memory Map and Register Descriptions
100XHD —100BASE-TX Half-Duplex
1 = Indicates the PHY supports 100BASE-TX half-duplex mode
0 = Indicates the PHY does not support 100BASE-TX half-duplex mode
10TFD —10BASE-T Full-Duplex
1 = Indicates the PHY supports 10BASE-T full-duplex mode
0 = Indicates the PHY does not support 10BASE-T full-duplex mode
10THD —10BASE-T Half-Duplex
1 = Indicates the PHY supports 10BASE-T half-duplex mode
0 = Indicates the PHY does not support 10BASE-T half-duplex mode
SUPPRE —MF Preamble Suppression
1 = Indicates that management frames are not required to contain the preamble stream
0 = Indicates that management frames are required to contain the preamble stream
ANCOMP —Auto-Negotiation Complete
To inform the management interface (MI) that it has completed processing, ANCOMP is set by the
A/N process. After it has been started, the auto-negotiation process uses link code words to exchange capability information and establish the highest common denominator (HCD) for link transactions.
1 = Indicates that the auto-negotiation process has completed and that the contents of registers 4 through 7 are valid.
0 = Indicates that the auto-negotiation process has not completed and that the contents of registers
4 through 7 are not valid
REMFLT — Remote Fault
Possible remote faults (RF) a) The link partner transmits the RF bit (5.13=1) b) Link partner protocol is not 00001 (5.4:0) c) Link partner advertises only T4 capability (5.9:5) d) No common operation mode found between PHY and the link partner.
After it is set, REMFLT is cleared each time register 1 is read via the management interface. REMFLT is also cleared by a PHY reset.
1 = Indicates that a remote fault condition has been detected.
0 = No fault detected
ANABL — Auto-Negotiation Ability
1 = Indicates that PHY has auto-negotiation ability
0 = Indicates that PHY does not have auto-negotiation ability
LNKSTST — Link Status
The PHY sets this bit when it determines that a valid link has been established. The occurrence of a link failure will cause LNKSTST to be cleared. After it has been cleared, it remains cleared until it is read via the management interface.
1 = Indicates a valid link has been established
0 = Indicates a valid link has NOT been established
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
19-11
Ethernet Physical Transceiver (EPHY) Block Description
JABDT —Jabber Detect
After it is set, JABDT is cleared each time register 1 is read via the management interface. JABDT is also cleared by a PHY reset. For 100BASE-TX operation, this signal will always be cleared.
1 = Indicates that a jabber condition has been detected
0 = Indicates that no jabber condition has been detected
EXCAP — Extended capability
1 = Indicates that the extended register set (registers 2–31) has been implemented in the PHY.
0 = Indicates that the extended register set (registers 2–31) has NOT been implemented in the PHY
19.3.3.3
EPHY Identifier Register 1
Registers $_02 and $_03 provide the PHY identification code.
MII Register Address 2 ($0x0002)
15
R
W
RESET: 0
14
0
13
0
12
0
11
01
10
0
9
0
8
PHYID
7
0 0
6
0
= Unimplemented or Reserved
Figure 19-8. EPHY Identifier Register 1
Read: Anytime
Write: Writes have no effect — Read only
PHYID — PHY ID Number
Composed of bits 3:18 of the organization unique identifier (OUI).
5
1
4
0
3
1
2
1
1
0
0
0
19.3.3.4
EPHY Identifier Register 2
Registers $_02 and $_03 provide the PHY identification code.
MII Register Address 3 ($0x0003)
15 14 13
PHYID
12
R
W
RESET: 0
11 10 9 8 7 6
MODELNUMBER
5
0 0 0 0
= Unimplemented or Reserved
0 0 0 0 0
Figure 19-9. EPHY Identifier Register 2
0
4
1
Read: Anytime
Write: Writes have no effect — Read only
PHYID — PHY ID number organization unique identifier. Composed of bits 15:10.
MODELNUMBER — Manufacturers model number. Composed of bits 9:4.
REVISIONNUMBER — Manufacturers revision number. Composed of bits 3:0.
3 2 1
REVISIONNUMBER
0
0 0 0 1
19-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Memory Map and Register Descriptions
19.3.3.5
Auto-Negotiate (A/N) Advertisement Register
The auto-negotiation (A/N) process requires four registers to communicate link information with its link partner: A/N advertisement register (MII register 4), A/N link partner ability register (MII register 5), A/N expansion register (MII register 6), and the A/N next page transmit register (MII register 7).
shows the contents of the A/N advertisement register. On power-up, before A/N starts, the register sets the selector field, bits 4.4:0, to 00001 to indicate that it is IEEE Standard 802.3 compliant. The technology ability fields (4.9:5) are set according to the values in the MII status register (1.15:11). The MI can set the technology ability field bits before renegotiations to allow management to auto-negotiate to an alternate common mode.
MII Register Address 4 ($0x0004)
R
W
RESET:
15
NXTP
1
14
0
0
13
RFLT
0
12
0
0
11
0
0
10
FLCTL
0
9
0
0
8
TAF
100FD
1
7
TAF
100HD
1
6
TAF
10FD
1
5
TAF
10HD
1
4
0
3 2 1
SELECTORFIELD[4:0]
0 0 0
0
1
= Unimplemented or Reserved
Figure 19-10. Auto Negotiate Advertisement Register
Read: Anytime
Write: Never
NXTP — Next Page
1 = Capable of sending next pages
0 = Not capable of sending next pages
RFLT — Remote Fault
1 = Remote fault
0 = No remote fault
FLCTL — Flow Control
1 = Advertise implementation of the optional MAC control sublayer and pause function as specified in IEEE standard clause 31 and anex 31B of 802.3. Setting FLCTL has no effect except to set the corresponding bit in the FLP stream
0 = No MAC-based flow control
TAF100FD — 100BASE-TX Full-Duplex
1 = 100BASE-TX full -duplex capable
0 = Not 100BASE-TX full-duplex capable
TAF100HD — 100BASE-TX Half-Duplex
1 = 100BASE-TX half-duplex capable
0 = Not 100BASE-TX half-duplex capable
TAF10FD — 10BASE-T Full-Duplex
1 = 10BASE-T full-duplex capable
0 = Not 10BASE-T full-duplex capable
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
19-13
Ethernet Physical Transceiver (EPHY) Block Description
TAF10HD — 10BASE-T Half-Duplex
1 = 10BASE-T half-duplex capable
0 = Not 10BASE-T half-duplex capable
19.3.3.6
Auto Negotiation Link Partner Ability (Base Page)
the MI and will be written by the auto-negotiation process when it receives a link code word advertising the capabilities of the link partner. This register has a dual purpose: exchange of base page information as shown in
Figure 19-11 , and exchange of next page information as shown in
.
MII Register Address 5 ($0x0005) (Base Page)
15 14 13 12 11 10
R
W
RESET:
NXTP
X
ACK RFLT TAF[1:0] FCTL
X X X X
= Unimplemented or Reserved
X
9
TAF
100T4
X
8
TAF
100FD
7
TAF
100HD
X X
6
TAF
10FD
X
5
TAF
10HD
X
4
X
3 2 1
SELECTORFIELD[4:0]
X
Figure 19-11. Auto Negotiation Link Partner Ability Register (Base Page)
X X
0
X
Read:
Write:
NXTP — Next Page
1 = Link partner capable of sending next pages
0 = Link partner not capable of sending next pages
ACK — Acknowledge
1 = Link Partner has received link code word
0 = Link Partner has not received link code word
RFLT — Remote Fault
1 = Remote fault
0 = No remote fault
FLCTL — Flow Control
1 = Advertises implementation of the optional MAC control sublayer and pause function as specified in IEEE standard clause 31 and anex 31B of 802.3. Setting FLCTL has no effect on the PHY.
0 = No MAC-based flow control
TAF100T4 — 100BASE-T4 Full-Duplex
1 = Link partner is 100BASE-T4 capable
0 = Link partner is not 100BASE-T4 capable
This function is not implemented in the EPHY.
TAF100FD — 100BASE-TX Full-Duplex
1 = Link partner is 100BASE-TX full-duplex capable
0 = Link partner is not 100BASE-TX full-duplex capable
19-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Memory Map and Register Descriptions
TAF100HD — 100BASE-TX Half-Duplex
1 = Link partner is 100BASE-TX half-duplex capable
0 = Link partner is not 100BASE-TX half-duplex capable
TAF10FD — 10BASE-T Full-Duplex
1 = Link partner is10BASE-T full-duplex capable
0 = Link partner is not 10BASE-T full-duplex capable
TAF10HD — 10BASE-T Half-Duplex
1 = Link partner is 10BASE-T half-duplex capable
0 = Link partner is not 10BASE-T half-duplex capable
19.3.3.7
Auto Negotiation Link Partner Ability (Next Page)
MII Register Address 5 ($0x0005) (Next Page)
R
W
RESET:
15
NXTP
X
14 13 12
ACK MSGP ACK2
11
TGL
10
X X X X
= Unimplemented or Reserved
X
9
X
8
X
7 6 5 4 3
Message/Unformatted Code Field [10:0]
X X X X X
Figure 19-12. Auto Negotiation Link Partner Ability Register (Next Page)
2
X
1
X
0
X
Read: Anytime
Write: See each field description
NXTP — Next Page
1 = Additional next pages will follow
0 = Last page transmitted
ACK — Acknowledge
ACK is used to acknowledge receipt of information.
1 = Link partner has received link code word
0 = Link partner has not received link code word
MSGP — Message Page
1 = Message page
0 = Unformatted page
ACK2 — Acknowledge 2
ACK2 is used to indicate that the receiver is able to act on the information (or perform the task) defined in the message.
1 = Receiver is able to perform the task defined in the message
0 = Receiver is unable to perform the task defined in the message
TGL — Toggle
1 = Previous value of the transmitted link code word equalled 0
0 = Previous value of the transmitted link code word equalled 1
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
19-15
Ethernet Physical Transceiver (EPHY) Block Description
Message/Unformatted Code Field
Message code field — Predefined code fields defined in IEEE 802.3u-1995 Annex 28C
Unformatted code filed — 11-bit field containing an arbitrary value
19.3.3.8
Auto-Negotiation Expansion Register
shows the contents of the A/N expansion register. The MI process can only read this register.
This register contains information about the A/N capabilities of the port’s link partner and information on the status of the parallel detection mechanism.
MII Register Address 6 ($0x0006)
R
W
RESET:
15
0
0
14
0
13
0
12
0
11
0
10
0
0 0 0 0
= Unimplemented or Reserved
0
9
0
0
8
0
0
7
0
0
6
0
0
5
0
0
4 3 2 1 0
PDFLT LPNPA NXTPA PRCVD LPANA
Figure 19-13. Auto-Negotiation Expansion Register
0 0 1 0 0
Read: Anytime
Write: Never
PDFLT — Parallel Detection Fault
This bit is used to indicate that zero or more than one of the NLP receive link integrity test function for
100BASE-TX have indicated that the link is ready (link_status=READY) when the A/N wait timer has expired. PDFLT will be reset to 0 after a read of register 6.
1 = Parallel detection fault has occurred
0 = Parallel detection fault has not occurred
LPNPA — Link Partner Next Page Able
Bit to indicate whether the link partner has the capability of using NP.
1 = Link partner is next page able
0 = Link partner is not next page able
NXTPA — Next Page Able
This bit is used to inform the MI and the link partner whether the port has next page capabilities.
1 = The port has next page capabilities
0 = The port does not have next page capabilities
PRCVD — Page Received
Bit is used to indicate whether a new link code word has been received and stored in the A/N link partner ability register (MII register 5). PRCVD is reset to 0 after register 6 is read.
1 = Three identical and consecutive link code words have been received from link partner
0 = Three identical and consecutive link code words have not been received from link partner
19-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Memory Map and Register Descriptions
LPANA — Link Partner A/N Able
Indicates whether the link partner has A/N capabilities.
1 = Link partner is A/N able
0 = Link partner is not A/N able
19.3.3.9
Auto Negotiation Next Page Transmit
shows the contents of the A/N next page transmit register. The MI writes to this register if it needs to exchange more information with the link partner. The PHY defaults to sending only a NULL message page to the link partner unless the STA overrides the values in the register. Next pages will be transmitted until the link partner has no more pages to transmit and bit 7.15 has been cleared by the STA.
MII Register Address 7 ($0x0007)
15 14
0
13 12
R
W
RESET:
NXTP
0 0
MSGP ACK2
1 0
11
TGL
0
10
0
9
0
8
0
7 6 5 4 3
Message/Unformatted Code Field [10:0]
0 0 0 0 0
2
0
1
0
0
1
= Unimplemented or Reserved
Figure 19-14. Auto Negotiation Next Page Transmit Register
Read: Anytime
Write: Never
NXTP — Next Page
1 = Additional next pages will follow
0 = Last page to transmit
MSGP — Message Page
1 = Message page
0 = Unformatted page
ACK2 — Acknowledge 2
ACK2 is used to indicate that the receiver is able to act on the information (or perform the task) defined in the message.
1 = Receiver is able to perform the task defined in the message
0 = Receiver is unable to perform the task defined in the message
TGL — Toggle
1 = Previous value of the transmitted link code word equalled 0
0 = Previous value of the transmitted link code word equalled 1
Message/Unformatted Code Field
Message code field — Predefined code fields defined in IEEE 802.3u-1995 Annex 28C
Unformatted code field — Eleven bit field containing an arbitrary value
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
19-17
Ethernet Physical Transceiver (EPHY) Block Description
19.3.4
PHY-Specific Registers
PHY also contains a number of registers to set its internal mode of operation. These registers can be set through the external management interface to determine capabilities such as speed, test-mode, circuit bypass mode, interrupt setting, etc. The PHY register set includes registers 16 through 29. These registers are not part of the MCU memory map.
19.3.4.1
Interrupt Control Register
MII Register Address 16 ($0x0010)
15
0
14 13 12
R
W
RESET: 0
ACKIE PRIE
0 0
LCIE
0
11
ANIE PDFIE RFIE JABIE
0
10
0
9
0
8
0
7
0
0
6 5
ACKR PGR
4
LKC
3
ANC
2
1 0
RMTF JABI
0 0 0 0 0 0 0
= Unimplemented or Reserved
Figure 19-15. Interrupt Control Register
Read: Anytime
Write: Anytime
ACKIE — Acknowledge Bit Received Interrupt Enable
1 = Enable interrupt when the acknowledge bit is received from the link partner
0 = Disable interrupt when acknowledge bit is received
PRIE — Page Received INT Enable
1 = Enable interrupt when a new page is received
0 = Disable interrupt when a page is received
LCIE — Link Changed Enable
1 = Enable interrupt when the link status changes
0 = Disable interrupt when the link status changes
ANIE — Auto-Negotiation Changed Enable
1 = Enable interrupt when the state of the auto-negotiation state machine has changed since the last access of this register
0 = Disable interrupt when the state of the auto-negotiation state machine has changed since the last access of this register
PDFIE — Parallel Detect Fault Enable
1 = Enable interrupt on a parallel detect fault
0 = Disable interrupt on a parallel detect fault
RFIE — Remote Fault Interrupt Enable
1 = Enable interrupt on a parallel detect fault
0 = Disable interrupt on a parallel detect fault
JABIE — Jabber Interrupt Enable
1 = Enable setting interrupt on detection of a jabber condition
0 = Disable setting interrupt on detection of a jabber condition
19-18
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Memory Map and Register Descriptions
ACKR — Acknowledge Bit Received
1 = Acknowledge bit has been received from the link partner
0 = Acknowledge bit has not been received since the last access of this register. (ACK bit 14 of the auto-negotiation link partner ability register was set by receipt of link code word)
PGR — Page Received
1 = A new page has been received from the link partner
0 = A new page has not been received from the link partner since the last access of this register (Bit
1 was set by a page received event)
LKC — Link Changed
1 = The link status has changed since the last access of this register
0 = The link status has not changed since the last access of this register. (LNK bit 14 of the proprietary status register was changed)
ANC — Auto-Negotiation Changed
1 = The auto-negotiation status has changed since the last access of this register
0 = The auto-negotiation status has not changed since the last access of this register
PDF — Parallel Detect Fault
1 = A parallel-detect fault has occurred since the last access of this register
0 = A parallel-detect fault has not been detected since the last access of this register. (Bit 4 was set by rising edge of parallel detection fault)
RMTF — Remote Fault
1 = A remote fault condition has been detected since the last access of this register
0 = A remote fault condition has not been detected since the last access of this register. (RMTF bit 4 of the status register was set by rising edge of a remote fault)
JABI — Jabber Interrupt
1 = A jabber condition has been detected since the last access of this register
0 = A jabber condition has not been detected since the last access of this register (JABD bit 1 of the status register was set by rising edge of jabber condition)
19.3.4.2
Proprietary Status Register
MII Register Address 17 ($0x0011)
15 14 13 12 11
R
W
RESET:
0
0
LNK DPMD SPD 0
10 9
ANNC PRCVD
8
ANC
MODE
7
0
6
0
5
PLR
1 1 1 0
= Unimplemented or Reserved
0 0 (1) 0 0 0
Figure 19-16. Proprietary Status Register
Read: Anytime
Write:
4
0
0
3
0
0
2
0
0
1
0
0
0
0
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
19-19
Ethernet Physical Transceiver (EPHY) Block Description
LNK — Link Status
This is a duplicate of LNKSTAT bit 2 of the status register (1.2).
1 = Link is down
0 = Link is up
DPMD — Duplex Mode
1 = Full-duplex
0 = Half-duplex
SPD — Speed
1 = 100 Mbps
0 = 10 Mbps
ANNC — Auto-Negotiation Complete
This is a duplicate of ANCOMP bit 5 of the status register (1.5)
1 = A-N complete
0 = A-N not complete
PRCVD — Page Received
1 = Three identical and consecutive link code words have been received
0 = Three identical and consecutive link code words have not been received
ANCMODE — Auto-Negotiation (A-N) Common Operating Mode
This bit is only valid while the ANNC bit 10 is 1
1 = A common operation mode was not found
0 = A-N is complete and a common operation mode has been found
PLR — Polarity Reversed (10BASE-T)
1 = 10BASE-T receive polarity is reversed
0 = 10BASE-T receive polarity is normal
19.3.4.3
Proprietary Control Register
MII Register Address 18 ($0x0012)
R
W
RESET:
15
0
0
14
FE
FLTD
0
13
MIILBD
1
12
0
0
11
1
1
10
JBDE
1
9
LNK
TSTD
0
8
POL
CORD
0
7
ALGD
0
6
ENC
BYP
0
5
SCR
BYP
0
4
TRD
ANALB
0
3
TR
TST
0
2
0
0
= Unimplemented or Reserved
Figure 19-17. Proprietary Control Register
The miscellaneous (EMISC) register provides visibility of internal counters used by the EMAC.
Read: Anytime
Write: Anytime
1
0
0
0
0
0
19-20
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Functional Description
FEFLTD — Far End Fault Disable
1 = Far end fault detect is disabled
0 = Far end fault detect on receive and transmit is enabled. This applies only while auto-negotiation is disabled
MIILBO — MII Loopback Disable
1 = Disable MII loopback
0 = MII transmit data is looped back to the MII receive pins
JBDE — Jabber Detect Enable (10BASE-T)
1 = Enable jabber detection
0 = Disable jabber detection
LNKTSTD — Link Test Disable (10BASE-T)
1 = Disable 10BASE-T link integrity test
0 = 10BASE-T link integrity test enabled
POLCORD — Disable Polarity Correction (10BASE-T)
1 = 10BASE-T receive polarity correction is disabled
0 = 10BASE-T receive polarity is automatically corrected
ALGD — Disable Alignment
1 = Un-aligned mode. Available only in symbol mode
0 = Aligned mode
ENCBYP — Encoder Bypass
1 = Symbol mode and bypass 4B/5B encoder and decoder
0 = Normal mode
SCRBYP — Scrambler Bypass Mode (100BASE-TX)
1 = Bypass the scrambler and de-scrambler
0 = Normal
TRDANALB — Transmit and Receive Disconnect and Analog Loopback
1 = High-impedance twisted pair transmitter. Analog loopback mode overrides and forces this bit
0 = Normal operation
TRTST — Transmit and Receive Test (100BASE-TX)
1 = Transmit and receive data regardless of link status
0 = Normal operation
19.4
Functional Description
The Ethernet physical interface is an IEEE 802.3 compliant 10/100 Ethernet physical transceiver. The
Ethernet physical interface can be configured to support 10BASE-T or 100BASE-TX applications. The
Ethernet physical interface is configurable via internal registers which are accessible through the MII management interface as well as limited configurability using the EPHY register map.
There are five basic modes of operation for the EPHY:
• Power down/initialization
• Auto-negotiate
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
19-21
Ethernet Physical Transceiver (EPHY) Block Description
• 10BASE-T
• 100BASE-TX
• Low-power
19.4.1
Power Down/Initialization
Upon reset, the EPHYEN bit, in the Ethernet physical transceiver control register 0 (EPHYCTL0), is cleared and EPHY is in its lowest power consumption state. All analog circuits are powered down. The twisted-pair transmitter and receiver pins (PHY_TXP, PHY_TXN, PHY_RXP, and PHY_RXN) are high-impedance. The MII management interface is not accessible. All MII registers are initialized to their reset state. The ANDIS, DIS100, and DIS10 bits, in the EPHYCTL0 register, have no effect until the
EPHYEN bit is set.
The EPHYEN bit can be set or cleared by a register write at any time. Prior to enabling the EPHY, setting
EPHYEN to 1, the MII PHY address PHYADD[4:0] must be set in the Ethernet physical transceiver control register 1 (EPHYCTL1), and the ANDIS, DIS100, DIS10 bits, in the EPHYCTL0 register, must be configured for the desired start-up operation. Whenever the EPHYEN bit transitions from 0 to 1, MDIO communications must be delayed until the completion of a start-up delay period (t
Start-up
).
, see
19-22
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
RESET or EPHYEN=0
Set PHYADD[4:0], and
ANDIS, DIS100, DIS10
Set EPHYEN=1
PHYADD[4:0] and ANDIS become latched in MII registers
Delay for t
Start-up
Configure MII registers via MDIO
Initialization Complete
Figure 19-18. EPHY Start-Up / Initialization Sequence
Functional Description
EPHYEN
MDIO t
Start-up
Figure 19-19. EPHY Start-Up Delay
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
19-23
Ethernet Physical Transceiver (EPHY) Block Description
If the auto-negotiation mode of operation is desired, the ANDIS bit in the EPHYCTL0 must be set to 0 and the DIS100 and DIS10 bits must be cleared prior to setting EPHYEN to 1. Refer to
for more information on auto-negotiation operation.
If the mode of operation will be set manually, the ANDIS bit must be set to 1 in the EPHYCTL0 register and the DIS100 and DIS10 bits must be cleared prior to setting EPHYEN to 1. After the EPHYEN bit has been set and the start-up delay period is completed, the mode of operation can be configured through the
MII registers. Table 19-3 summarizes the MII register configuration and operational modes.
Table 19-3. Operational Configuration While Auto-Negotiation is Disabled
1
Bit 0.12
Auto
Neg.
0
0
0
0
0
0
0
Bit 0.13
Data
Rate
1
1
0
0
1
1
1
Bit 0.8
Duplex
1
1
1
0
1
1
1
0 1 0
1
Symbol mode is not supported.
Bit 18.6
Encoder
Bypass
0
1
X
X
Bit 18.5
Scrambler
Bypass
0
0
X
X
1
1
1
0
0
1
1
0
Bit 18.7
Symbol
Unalign
0
0
X
X
1
0
1
0
Operation
10BASE-T full-duplex
10BASE-T half-duplex
100BASE-TX full-duplex
100BASE-TX full-duplex with encoder bypass (symbol mode) — aligned
100BASE-TX full-duplex with encoder bypass (symbol mode) — unaligned
100BASE-TX full-duplex with scrambler and encoder bypassed (symbol mode), aligned
100BASE-TX full-duplex with scrambler and encoder bypassed (symbol mode), unaligned
100BASE-TX half-duplex
19.4.2
Auto-Negotiation
Auto-negotiation is used to determine the capabilities of the link partner. Auto-negotiation is compliant with IEEE 802.3 clause 28. In this case, the PHY will transmit fast link pulse (FLP) bursts to share its capabilities with the link partner.
If the link partner is also capable of performing auto-negotiation, it will also send FLP bursts. The information shared through the FLP bursts will allow both link partners to find the highest common mode
(if it exists).
If no common mode is found, the remote fault bit (1.4) will be set. A remote fault is defined as a condition in which the PHY and the link partner cannot establish a common operating mode. Configuring auto-negotiation advertisement register sets the different auto-negotiation advertisement modes.
If the link partner does not support auto-negotiation, it will transmit either normal link pulses (NLP) for
10 Mbps operation, or 100 Mbps idle symbols. Based on the received signal, the PHY determines whether the link partner is 10 Mbps capable or 100 Mbps capable. The ability to do this is called parallel detection.
If using parallel detection, the link will be configured as a half-duplex link. After parallel detection has established the link configuration, the remote fault bit will be set if the operating mode does not match the pre-set operating modes.
19-24
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Functional Description
shows the main blocks used in the auto-negotiation function. The transmit block allows transmission of fast link pulses to establish communications with partners that are auto-negotiation able.
The receive block determines the capabilities of the link partner and writes to the link partner ability register (register 5). The arbitration block determines the highest common mode of operation to establish the link.
MR_ADV_ABILITY[15:0] MR_PARALLEL_DETECTION_FAULT
MR_AUTONEG_COMPLETE
MR_ADV_ABILITY[3:0]
MR_LP_ADV_ABILITY[15:0]
MR_LP_ADV_ABILITY[3:0]
CLK
POWER_ON
MR_MAIN_RESET
MR_AUTONEG_ENABLE
TRANSMIT complete_ack transmit_ability transmit_ack transmit_disable flp_link_good ack_finished
ARBITRATION flp_receive_idle match_wo_ack match_w_ack receive_done flp_link_good
RECEIVE
1
2
1 LP_LINK_CONTROL[1:0]
2 LP_LINK_STATUS[1:0]
NLP RECEIVE LINK
INTEGRITY TEST
TD_AUTONEG TX_LINK_CONTROL[1:0]
TX_LINK_STATUS[1:0]
DO RD
LINK_TEST_RECEIVE
Figure 19-20. Auto-Negotiation
LINKPULSE
19.4.3
10BASE-T
The 10BASE-T interface implements the physical layer specification for a 10 Mbps over two pairs of twisted-pair cables. The specifications are given in clause 14 of the IEEE 802.3 standard.
In 10BASE-T mode, Manchester encoding is used. When transmitting, nibbles from the MII are converted to a serial bit stream and then Manchester encoded. When receiving, the Manchester encoded bit stream is decoded and converted to nibbles for presentation to the MII.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
19-25
Ethernet Physical Transceiver (EPHY) Block Description
A 2.5 MHz internal clock is used for nibble wide transactions. A 10 MHz internal clock is used for serial transactions.
TX
MII
PARALLEL
TO
SERIAL
MANCHESTER
ENCODER
DIGITAL
FILTER
PHY_TXN
PHY_TXP
CARRIER
SENSE JABBER
DIGITAL
LOOPBACK
(bit 0.14)
LINE TRANSMITTER/
LINE RECEIVER
RX
MII
SERIAL
TO
PARALLEL
MANCHESTER
DECODER
AND
TIMING
RECOVERY
POLARITY
CHECK
SQUELCH
Figure 19-21. 10BASE-T Block Diagram
PHY_RXN
PHY_RXP
Parallel to Serial: Converts the 4-bit wide nibbles from the MII to serial format before the information is processed by subsequent blocks.
Manchester Encoder: Allows encoding of both the clock and data in one bit stream. A logical one is encoded as a zero when the clock is high and a one when the clock is low. A logical zero is encoded as a one when the clock is high and a zero when the clock is low.
Digital Filter: Performs pre-emphasis and low pass filtering of the input Manchester data.
DAC: Converts the digital data to an analog format before transmission on the media.
Carrier Sense: In half-duplex operation, carrier is asserted when either the transmit or receive medium is active. In full-duplex operation, carrier asserted only on reception of data. During receive, carrier sense is asserted during reception of a valid preamble, and de-asserted after reception of an EOF.
Loopback: Enabled when bit 0.14 is asserted. This loopback mode allows for the Manchester encoded and filtered data to be looped back to the squelch block in the receive path. All the 10BASE-T digital functions are exercised during this mode. The transmit and receive channels are disconnected from the media.
MII loopback (18.13) must be disabled to allow for correct operation of the digital loopback (0.14).
Link Generator: Generates a 100 ns duration pulse at the end of every 12 ms period of the transmission path being idle (TXEN de-asserted). This pulse is used to keep the 10BASE-T link operational in the absence of data transmission.
19-26
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Functional Description
Link Integrity Test: Used to determine whether the 10BASE-T link is operational. If neither data nor a link pulse is received for 64 ms, then the link is considered down. While the link is down, the transmit, loopback, collision detect, and SQE functions are disabled. The link down state is exited after receiving data or four link pulses.
Jabber: Prevents the transmitter from erroneously transmitting for too long a period. The maximum time the device can transmit is 50,000 bit times. When the jabber timer is exceeded, the transmit output goes idle for 0.525 s.
This function can be disabled with the jabber inhibit register bit (18.10).
Squelch: Used to determine whether active data, a link pulse, or an idle condition exists on the 10BASE-T receive channel. While an idle or link pulse condition exists, a higher squelch level is used for greater noise immunity. The squelch output is used to determine when the Manchester decoder should operate. The output is also used to determine when an end of packet is received.
Polarity Check: By examining the polarity of the received link pulses, EPHY can determine whether the received signal is inverted. If the pulses are inverted, this function changes the polarity of the signal.This feature is activated if eight inverted link pulses are received or four frames with inverted EOF are encountered.
Manchester Decoder and Timing Recovery: Decodes the Manchester encoded data. The receive data and clock are recovered during this process.
Serial to Parallel: Converts the serial bit stream from the Manchester decoder to the required MII parallel format.
PMD Sublayer: Transmits and receives signals compliant with IEEE 802.3, Section 14.
Line Transmitter and Line Receiver: These analog blocks allow the EPHY to drive and receive data from the 10BASE-T media.
19.4.4
100BASE-TX
100BASE-TX specifies operation over two pairs of category 5 unshielded twisted-pair cable (UTP).
The EPHY implementation includes the physical coding sublayer (PCS), the physical medium attachment
(PMA), and the physical medium dependent (PMD) sublayer.
The block diagram for 100BASE-TX operation is shown in Figure 19-22 .
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
19-27
Ethernet Physical Transceiver (EPHY) Block Description
PMA
TX
PCS
4B5B
ENCODER
PARALLEL
TO
SERIAL
SCRAMBLER
PMD
MLT3
ENCODER
SLOPE
CONTROL
MII
CARRIER
SENSE
PMA
LINE
DRIVER
DIGITAL
LOOPBACK
(bit 0.14)
ANALOG
LOOPBACK
(bit 18.4)
LINK
MONITOR
RX
4B5B
DECODER
SERIAL TO
PARALLEL
AND SYMBOL
ALIGNMENT
DESCRAMBLER
MLT3
DECODER
EQUALIZER
AND
TIMING
RECOVERY
BASELINE
WANDER
Figure 19-22. 100BASE-TX Block Diagram
19.4.4.1
Sublayers
19.4.4.1.1
PCS Sublayer
The PCS sublayer is the MII interface that provides a uniform interface to the reconciliation sublayer.
The services provided by the PCS include:
• Encoding/decoding of MII data nibbles to/from 5-bit code-groups (4B/5B)
• Carrier sense and collision indications
• Serialization/deserialization of code-groups for transmission/reception on the PMA
• Mapping of transmit, receive, carrier sense, and collision detection between the MII and the underlying PMA
Serial to Parallel and Symbol Alignment: This block looks for the occurrence of the JK symbol to align the serial bit stream and convert it to a parallel format.
Carrier Sense: In full-duplex mode, carrier sense is only asserted while the receive channel is active . The carrier sense examines the received data bit stream looking for the SSD, the JK symbol pair. In the idle state, IDLE symbols (all logic ones) will be received. If the first 5-bit symbols received after an idle stream forms the J symbol (11000) it asserts the CRS signal. At this point the second symbol is checked to confirm the K symbol (10001). If successful, the following aligned data (symbols) are presented to the 4B/5B decoder. If the JK pair is not confirmed, the false carrier detect is asserted and the idle state is re-entered.
Carrier sense is de-asserted when the ESD (end-of-stream) delimiter, the TR symbol pair, is found, or when an idle state is detected.
In half-duplex, CRS is also asserted on transmit.
Parallel to Serial: This block takes parallel data and converts it to serial format.
19-28
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Functional Description
4B/5B Encoder/Decoder : The 4B/5B encoder converts the 4-bit nibbles from the reconciliation sublayer to a 5-bit code group.
19.4.4.1.2
PMA Sublayer
The PMA provides medium-independent means for the PCS and other bit-oriented clients (e.g., repeaters) to support the use of a range of physical media. For 100BASE-TX the PMA performs these functions:
• Mapping of transmit and receive code-bits between the PMA’s client and the underlying PMD
• Generating a control signal indicating the availability of the PMD to a PCS or other client
• Synchronization with the auto-negotiation function
• Generating indications of carrier activity and carrier errors from the PMD
• Recovery of clock from the NRZI data supplied by the PMD
19.4.4.1.3
PMD Sublayer
For 100BASE-TX, the ANSI X3.263: 199X (TP-PMD) standard is used. These signalling standards, called PMD sublayers, define 125 Mbps, full-duplex signalling systems that use STP and UTP wiring.
Scrambler/De-scrambler: The scrambler and de-scrambler used in EPHY meet the ANSI Standard
X3.263-1995 FDDI TP-PMD. The purpose of the scrambler is to randomize the 125 Mbps data on transmission resulting in a reduction of the peak amplitudes in the frequency spectrum. The de-scrambler restores the received 5-bit code groups to their unscrambled values.
The scrambler input data (plaintext) is encoded by modulo 2 addition of a key stream to produce a ciphertext bit stream. The key stream is a periodic sequence of 2047 bits generated by the recursive linear function X[n] = X[n-11] + X[n-9] (modulo 2).
If not transmitting data, the scrambler encodes and transmits idles. This allows a pattern to use by the de-scrambler to synchronize and decode the scrambled data.
The implementation of the scrambler and de-scrambler is as shown in Appendix G of the ANSI Standard
X3.263-1995.
For test, the scrambler can be bypassed by setting bit 18.5. Scrambler bypass mode is a special type of interface for 100BASE-TX operation that bypasses the scrambler and de-scrambler operation. This mode is typically used for test so that input and output test vectors match. In this mode, idles are not sent and the
MAC must provide idles.
MLT-3 Encoder/Decoder: An MLT-3 encoder is used in the transmit path to convert NRZ bit stream data from the PMA sublayer into a three-level code. This encoding results in a reduction in the energy over the critical frequency range. The MLT-3 decoder converts the received three-level code back to an NRZ bit stream.
Baseline Wander: The use of the scrambler and MLT-3 encoding can cause long run lengths of 0s and 1s that can produce a DC component. The DC component cannot be transmitted through the isolation transformers and results in baseline wander. Baseline wander reduces noise immunity because the base line moves nearer to either the positive or negative signal comparators. To correct for this EPHY uses DC
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
19-29
Ethernet Physical Transceiver (EPHY) Block Description restoration to restore the lost DC component of the recovered digital data to correct the baseline wander problem.
Timing Recovery : The timing recovery block locks onto the incoming data stream, extracts the embedded clock, and presents the data synchronized to the recovered clock.
In the event that the receive path is unable to converge to the receive signal, it resets the MSE-good (bit
25.15) signal. The clock synthesizer provides a center frequency reference for operation of the clock recovery circuit in the absence of data.
Adaptive Equalizer: At a data rate of 125 Mbps, the cable introduces significant distortion due to high frequency roll off and phase shift. The high frequency loss is mainly due to skin effect, which causes the conductor resistance to rise as the square of the frequency.
The adaptive equalizer will compensate for signal amplitude and phase distortion incurred from transmitting with different cable lengths.
Loopback : If asserted by bit 0.14, data encoded by the MLT3 encoder block is looped back to the MLT3 decoder block while the transmit and receive paths are disconnected from the media.
A second loopback mode for 100BASE-TX is available by setting bit 18.13 (MII loopback) to a logical 1.
This loopback mode takes the MII transmit data and loops it directly back to the MII receive pins. Again, the transmit and receive paths are disconnected from the media.
MII loopback has precedence over the digital loopback if both are enabled at the same time.
A third loopback mode is available by setting bit 18.4 high. This analog loopback mode takes the MLT3 encoded data and loops it back through the base line wander and analog receive circuits.
Line Transmitter and Line Receiver: These analog blocks allow EPHY to drive and receive data to/from the 100BASE-TX media. The transmitter is designed to drive a 100Ω UTP cable.
Link Monitor : The link monitor process is responsible for determining whether the underlying receive channel is providing reliable data. If a failure is found, normal operation will be disabled. As specified in the IEEE 802.3 standard, the link is operating reliably if a signal is detected for a period of 330 µ s.
Far End Fault : While the auto-negotiation function is disabled, this function is used to exchange fault information between the PHY and the link partner.
19.4.5
Low Power Modes
There are several reduced power configurations available for the EPHY.
19.4.5.1
Stop Mode
If the MCU executes a STOP instruction, the EPHY will be powered down and all internal MII registers reset to their default state. Upon exiting stop mode, the EPHY will exit the power-down state and latch the values previously written to the EPHYCTL0 and EPHYCTL1 registers. The MII registers will have to be re-initialized after the start-up delay (t
Start-up
) has expired.
19-30
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Functional Description
19.4.5.2
Wait Mode
If the MCU executes a WAIT instruction with the EPHYWAI bit set, the EPHY will be powered down and all internal MII registers reset to their default state. Upon exiting STOP mode the EPHY will exit the power-down state and latch the values previously written to the EPHYCTL0 and EPHYCTL1 registers.
The MII registers must be re-initialized after the start-up delay (t
Start-up
) has expired.
19.4.5.3
MII Power Down
This mode disconnects the PHY from the network interface (three-state receiver and driver pins).
Setting bit 0.11 of the port enters this mode. In this mode, the management interface is accessible but all internal chip functions are in a zero power state.
In this mode all analog blocks except the PLL clock generator and band gap reference are in low power mode. All digital blocks except the MDIO interface and management registers are inactive.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
19-31
Ethernet Physical Transceiver (EPHY) Block Description
19-32
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 20
DMA Controller Module
20.1
Introduction
This chapter describes the direct memory access (DMA) controller module. It provides an overview of the module and describes in detail its signals and registers. The latter sections of this chapter describe operations, features, and supported data transfer modes in detail.
NOTE
The designation ‘ n ’ is used throughout this section to refer to registers or signals associated with one of the four identical DMA channels: DMA0,
DMA1, DMA2, or DMA3.
20.1.1
Overview
The DMA controller module enables fast transfers of data, providing an efficient way to move blocks of data with minimal processor interaction. The DMA module, shown in
, has four channels that allow byte, word, longword, or 16-byte burst data transfers. Each channel has a dedicated source address register (SAR n ), destination address register (DAR n ), byte count register (BCR n ), control register
(DCR n ), and status register (DSR n ). Transfers are dual address to on-chip devices, such as UART and
GPIOs.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 20-1
DMA Controller Module
MADDR[31:0]
MWDATA[31:0]
MRDATA[31:0]
MRSTH
MRW
MSIZ[1:0]
MTT[1:0]
MTM[2:0]
MTA
MTEA
MPL
#
#
#
MAP
MDP
MAH
* MTS
# - Version 3 Mbus signal
* - Version 2 Mbus signal
DMA
SADDR[4:2]
SDATA[31:0]
SSRSTH
SMEN[3:0]
SBE[3:0]
SOE
SWE
SIVOE
SINT[3:0]
EXT_REQ[3:0]
20-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
DREQ0 DREQ1 DREQ2 DREQ3
Internal
Bus
Channel
Requests
Channel 0 Channel 1 Channel 2 Channel 3
SAR0
DAR0
BCR0
DCR0
DSR0
SAR1
DAR1
BCR1
DCR1
DSR1
SAR2
DAR2
BCR2
DCR2
DSR2
SAR3
DAR3
BCR3
DCR3
DSR3
Channel
Attributes
Channel
Enables
MUX
Control
MUX
Interrupts
System Bus Address
System Bus Size
Current Master Attributes
DMA Controller Module
Read Data Bus
Data Path
Data Path
Control
Write Data Bus
Arbitration/
Control
Bus Interface
Registered
Bus Signals
Figure 20-1. DMA Signal Diagram
NOTE
Throughout this chapter, the terms ‘external request’ and ‘DREQ’ are used to refer to a DMA request from one of the on-chip UARTS, DMA timers.
For details on the connections associated with DMA request inputs, see
Section 20.4.1, “DMA Request Control (DMAREQC) .”
20.1.2
Features
The DMA controller module features the following:
• Four independently programmable DMA controller channels
• Auto-alignment feature for source or destination accesses
• Dual-address transfers
• External request per channel
• Channel arbitration on transfer boundaries
• Data transfers in 8-, 16-, 32-, or 128-bit blocks using a 16-byte buffer
• Continuous-mode or cycle-steal transfers
• Independent transfer widths for source and destination
• Independent source and destination address registers
• Modulo addressing on source and destination addresses
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 20-3
DMA Controller Module
• Automatic channel linking
20.2
M-bus Priority Level (MPL)
This output signal indicates to the bus arbiter that the DMA has been programmed to have priority on the transfer. This signal is a decode of the BWC bits. Depending on the M-bus arbiter architecture, this signal may be ignored.
20.3
DMA Transfer Overview
The DMA module can transfer data faster than the ColdFire core. The term “direct memory access” refers to a fast method of moving data within system memory (including memory and peripheral devices) with minimal processor intervention, greatly improving overall system performance. The DMA module consists of four independent, functionally equivalent channels, so references to DMA in this chapter apply to any of the channels. It is not possible to implicitly address all four channels at once.
The processor generates DMA requests internally by setting DCR[START]; the UART modules and DMA timers can generate a DMA request by asserting internal DREQ signals. The processor can program bus bandwidth for each channel. The channels support cycle-steal and continuous transfer modes; see
Section 20.5.1, “Transfer Requests (Cycle-Steal and Continuous Modes) .”
The DMA controller supports dual-address transfers. The DMA channels support up to 32 data bits.
• Dual-address transfers—A dual-address transfer consists of a read followed by a write and is initiated by an internal request using the START bit or by a peripheral DMA request. Two types of
transfer can occur: a read from a source device or a write to a destination device. See Figure 20-2
for more information.
Control and Data
Memory/
Peripheral
Memory/
Peripheral
Control and Data
Figure 20-2. Dual-Address Transfer
Any operation involving the DMA module follows the same three steps:
1. Channel initialization—Channel registers are loaded with control information, address pointers, and a byte-transfer count.
2. Data transfer—The DMA accepts requests for operand transfers and provides addressing and bus control for the transfers.
20-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
DMA Controller Module
3. Channel termination—Occurs after the operation is finished, either successfully or due to an error.
The channel indicates the operation status in the channel’s DSR, described in Section 20.4.4.1,
.”
20.4
Memory Map/Register Definition
registers during a DMA transfer can result in undefined operation. Table 20-1
shows the mapping of DMA controller registers.
Table 20-1. Memory Map for DMA Controller Module Registers
DMA
Channel
IPSBAR Offset [31:24] [23:16] [15:8]
—
0
1
0x00_0014
0x00_0100
0x00_0104
0x00_0108
0x00_010C
0x00_0110
0x00_0114
Status Register 0
(DSR0)
DMA Request Control Register (DMAREQC)
Source Address Register 0 (SAR0)
Destination Address Register 0 (DAR0)
Byte Count Register 0 (BCR0)
Control Register 0 (DCR0)
Source Address Register 1 (SAR1)
Destination Address Register 1 (DAR1)
0x00_0118
Status Register 1
(DSR1)
2
0x00_011C
0x00_0120
0x00_0124
0x00_0128
Status Register 2
(DSR2)
3
0x00_012C
0x00_0130
0x00_0134
0x00_0138
Status Register 3
(DSR3)
0x00_013C
1
Located within the SCM, but listed here for clarity.
Byte Count Register 1 (BCR1)
Control Register 1 (DCR1)
Source Address Register 2 (SAR2)
Destination Address Register 2 (DAR2)
Byte Count Register 2 (BCR2)
Control Register 2 (DCR2)
Source Address Register 3 (SAR3)
Destination Address Register 3 (DAR3)
Control Register 3 (DCR3)
[7:0]
20.4.1
DMA Request Control (DMAREQC)
The DMAREQC register provides a software-controlled connection matrix for DMA requests. It logically routes DMA requests from the DMA timers and UARTs to the four channels of the DMA controller.
Writing to this register determines the exact routing of the DMA request to the four channels of the DMA modules.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 20-5
DMA Controller Module
IPSBAR
Offset:
0x00_0014 (DMAREQC)
30 29 28 31
R
W
R 0
W
Reset 0
0
0
0
0
0
0
27 26
Off-Platform DMAREQ Control
0
0
0
0
25
0
0
15
R
W
Reset 0
14 13
DMAC3
0
12 11 10 9
DMAC2
24
0
0
8
23
0
0
7
22
0
0
21
0
0
6 5
DMAC1
20
0
0
4
0
3
0 0 0 0 0 0 0 0 0 0
Figure 20-3. DMA Request Control Register (DMAREQC)
0
19 18 17
DMAREQC_EXT
16
0 0 0 0
0
Access: read/write
0
2 1
DMAC0
0 0
0
0
0
IPSBAR
Offset
0x00_0014
31
R
W
R 0
W
Reset 0
30 29 28 27 26
Off-Platform DMAREQ Control
25
0
0
0
0
0
0
0
0
0
0
0
0
15
R
W
Reset 0
14 13
DMAC3
0
12 11 10 9
DMAC2
24
0
0
8
23
0
0
0
7
22
0
0
21
0
0
0 0
6 5
DMAC1
20
0
0
0
4
19 18 17
DMAREQC_EXT
16
0 0 0 0 0 0 0 0 0 0
Figure 20-4. DMA Request Control Register (DMAREQC)
0
0
3
0 0
2 1
DMAC0
0
—
0
0
0
0
Table 20-2. DMAREQC Field Description
Field Description
OFP_DMA
REQC
This 8-bit field is reserved for an SoC-dependent specification involving off-platform DMA requests.
—
31–16
—
Reserved. Should be cleared.
Reserved, should be cleared.
20-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
DMA Controller Module
Table 20-2. DMAREQC Field Description (continued)
Field
15–0
DMAC n
Description
DMA channel n . Each four bit field defines the logical connection between the DMA requesters and that DMA channel.There are ten possible requesters (4 DMA Timers and 6 UARTs). Any request can be routed to any of the DMA channels. Effectively, the DMAREQC provides a software-controlled routing matrix of the 10 DMA request signals to the 4 channels of the DMA module. DMAC3 controls DMA channel 3, DMAC2 controls DMA channel 2, etc.
0100 DMA Timer 0.
0101 DMA Timer 1.
0110 DMA Timer 2.
0111 DMA Timer 3.
1000 UART0 Receive.
1001 UART1 Receive.
1010 UART2 Receive.
1100 UART0 Transmit.
1101 UART1 Transmit.
1110 UART2 Transmit.
All other values are reserved and will not generate a DMA request.
20.4.2
Source Address Registers (SAR0–SAR3)
SAR n , shown in
, contains the address from which the DMA controller requests data.
Access: read/write IPSBAR
Offset:
0x00_0100 (DMA0)
0x00_0110 (DMA1)
0x00_0120 (DMA2)
0x00_0130 (DMA3)
30 29 31
R
W
Reset 0 0 0
28
0
27
0
26
0
25
0
24
SAR
23
0 0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
R
W
Reset 0
14
0
13
0
12 11 10 9 8
SAR
7 6 5 4
0 0 0 0 0 0 0 0
Figure 20-5. Source Address Registers (SAR n )
0
3
0
2
0
1
0
0
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 20-7
DMA Controller Module
IPSBAR
Offset
0x00_0100 (DMA0)
0x00_0110 (DMA1)
0x00_0120 (DMA2)
0x00_0130 (DMA3)
30 29 31
R
W
Reset 0 0 0
28
0
27 26 25 24
SAR
23 22 21 20 19 18 17
0 0 0 0 0 0 0 0 0 0 0
15
R
W
Reset 0
14 13 12 11 10 9 8
SAR
7 6 5 4 3 2
0 0 0 0 0 0 0
Figure 20-6.
0 0 0 0 0 0
NOTE
The backdoor enable bit must be set in the SCM RAMBAR as well as the secondary port valid bit in the core RAMBAR in order to enable backdoor accesses from the DMA to SRAM. See Section 10.5.2, “Memory Base
Address Register (RAMBAR),” and
Section 11.2.1, “SRAM Base Address
for more details.
1
0
16
0
0
0
20.4.3
Destination Address Registers (DAR0–DAR3)
DAR n , shown in
, holds the address to which the DMA controller sends data.
IPSBAR
Offset:
0x00_0104 (DMA0)
0x00_0114 (DMA1)
0x00_0124 (DMA2)
0x00_0134 (DMA3)
30 29 31
R
W
Reset 0 0 0
28
0
27
0
26
0
25
0
24
DAR
23
0 0
22
0
21
0
20
0
19
0
Access: read/write
18
0
17
0
16
0
15
R
W
Reset 0
14
0
13
0
12 11 10 9 8
DAR
7 6 5 4
0 0 0 0 0 0 0 0 0
Figure 20-7. Destination Address Registers (DAR n )
3
0
2
0
1
0
0
0
20-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
DMA Controller Module
IPSBAR
Offset
0x00_0104 (DMA0)
0x00_0114 (DMA1)
0x00_0124 (DMA2)
0x00_0134 (DMA3)
30 29 31
R
W
Reset 0 0
14
0
13 15
R
W
Reset 0 0 0
28
0
12
0
27
0
11
0
26 25
0
10
0
9
24
DAR
23
0
8
DAR
0
7
0 0 0
Figure 20-8.
0
22 21 20 19 18 17 16
0
6
0
0
5
0
0
4
0
0
3
0
0
2
0
0
1
0 0
0
0
20.4.4
Byte Count Registers (BCR
n
) and DMA Status Registers (DSR
n
)
BCR n
, shown in Figure 20-9 , contains the number of bytes yet to be transferred for a given block. BCR
n decrements on the successful completion of the address transfer of a write transfer. BCR n decrements by
1, 2, 4, or 16 for byte, word, longword, or line accesses, respectively.
Access: read/write IPSBAR
Offset:
0x00_0108 (DMA0)
0x00_0118 (DMA1)
0x00_0128 (DMA2)
0x00_0138 (DMA3)
30 29 31
R
W
Reset 0 0 0
28
DSR
27
0 0
26
0
25
0
24
0
23
0
22
0
21
0
20
BCR
19
0 0
18
0
17
0
16
0
15
R
W
Reset 0
14 13 12 11 10 9 8 7 6 5 4 3 2
BCR
0 0 0 0 0 0 0 0 0 0 0 0
Figure 20-9. Byte Count Registers (BCR n ) and Status Registers (DSR n )
0
1
0
0
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 20-9
DMA Controller Module
IPSBAR
Offset
0x00_0108 (DMA0)
0x00_0118 (DMA1)
0x00_0128 (DMA2)
0x00_0138 (DMA3)
30 29 31
R
W
Reset 0 0 0
28
DSR
27
0 0
26 25 24 23 22 21 20
BCR
19 18 17 16
0 0 0 0 0 0 0 0 0 0 0
15
R
W
Reset 0
14 13 12 11 10 9 8
BCR
7 6 5 4 3 2 1 0
0 0 0 0 0 0 0 0
Figure 20-10.
0 0 0 0 0 0 0
DSR n [DONE], shown in
Figure 20-11 , is set when the block transfer is complete.
When a transfer sequence is initiated and BCR n [BCR] is not a multiple of 16, 4, or 2 when the DMA is configured for line, longword, or word transfers, respectively, DSR n [CE] is set and no transfer occurs. See
Section 20.4.4.1, “DMA Status Registers (DSRn).”
20.4.4.1
DMA Status Registers (DSR n )
In response to an event, the DMA controller writes to the appropriate DSR n bit,
to DSR n [DONE] results in action.
Access: read/write IPSBAR
Offset:
(DSR n )
6 7
0 R
W
Reset: 0
CE
5
BES
4
BED
3
0
2
REQ
0 0 0 0
Figure 20-11. DMA Status Registers (DSR n )
0
1
BSY
0
0
DONE
0
20-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
DMA Controller Module
Field
7
6
CE
5
BES
4
BED
3
2
REQ
1
BSY
0
DONE
IPSBAR
Offset
7
0
6
CE R
W
Reset 0 0
5
BES
4
BED
0 0
Figure 20-12.
0
3
0
2
REQ
0
1 0
BSY DONE
0 0
Table 20-3. DSR n Field Descriptions
Description
Reserved, should be cleared.
Configuration error. Occurs when BCR, SAR, or DAR does not match the requested transfer size, or if BCR = 0 when the DMA receives a start condition. CE is cleared at hardware reset or by writing a 1 to DSR[DONE].
0 No configuration error exists.
1 A configuration error has occurred.
Bus error on source
0 No bus error occurred.
1 The DMA channel terminated with a bus error during the read portion of a transfer.
Bus error on destination
0 No bus error occurred.
1 The DMA channel terminated with a bus error during the write portion of a transfer.
Reserved, should be cleared.
Request
0 No request is pending or the channel is currently active. Cleared when the channel is selected.
1 The DMA channel has a transfer remaining and the channel is not selected.
Busy
0 DMA channel is inactive. Cleared when the DMA has finished the last transaction.
1 BSY is set the first time the channel is enabled after a transfer is initiated.
Transactions done. Set when all DMA controller transactions complete, as determined by transfer count or error conditions. When BCR reaches zero, DONE is set when the final transfer completes successfully. DONE can also be used to abort a transfer by resetting the status bits. When a transfer completes, software must clear DONE before reprogramming the DMA.
0 Writing or reading a 0 has no effect.
1 DMA transfer completed. Writing a 1 to this bit clears all DMA status bits and can be used in an interrupt handler to clear the DMA interrupt and error bits.
20.4.5
DMA Control Registers (DCR0–DCR3)
DCR n , shown in
, is used for configuring the DMA controller module.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 20-11
DMA Controller Module
IPSBAR
Offset:
0x00_010C (DMA0)
0x00_011C (DMA1)
0x00_012C (DMA2)
0x00_013C (DMA3)
31 30 29
R
W
INT
Reset 0
EEXT
0
CS
0
28
AA
0
15
R
W
Reset 0
14
SMOD
13
0 0
27
0
26
BWC
0
25
0
24
0
0
23
0
0
22
SINC
21
SSIZE
20
0 0 0
19
DINC
0
18
DSIZE
17
0 0
16
0
START
0
12 11 10
DMOD
9 8 7
D_REQ
6
0
5 4
LINKCC
0 0 0 0 0 0 0 0
Figure 20-13. DMA Control Registers (DCR n )
0
3
LCH1
2
0 0
Access: read/write
1
LCH2
0
0 0
IPSBAR
Offset
0x00_010C (DMA0)
0x00_011C (DMA1)
0x00_012C (DMA2)
0x00_013C (DMA3)
31 30 29 28
R INT EEXT CS AA
W
Reset 0 0 0 0
15
R
W
Reset 0
14
SMOD
13
0 0
12
0
Field
31
INT
30
EEXT
29
CS
27 26
BWC
25
0
11
0 0
10
DMOD
9
24
0
23
0
22
SINC
0
21 20
SSIZE
19
DINC
0 0 0
18 17
DSIZE
0 0
16
0
START
0 0
8
0
7 6
D_REQ 0
5 4
LINKCC
3
LCH1
2 1
LCH2
0
0 0 0 0 0
Figure 20-14.
0 0 0 0 0 0 0
Table 20-4. DCR n Field Descriptions
Description
Interrupt on completion of transfer. Determines whether an interrupt is generated by completing a transfer or by the occurrence of an error condition.
0 No interrupt is generated.
1 SINT asserts Internal interrupt signal is enabled.
Enable external request. Care should be taken because a collision can occur between the START bit and DREQ n when EEXT = 1.
0 External request is ignored.
1 Enables external request to initiate transfer. The internal request (initiated by setting the START bit) is always enabled.
Cycle steal.
0 DMA continuously makes read/write transfers until the BCR decrements to 0.
1 Forces a single read/write transfer per request.
20-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Field
28
AA
27–25
BWC
DMA Controller Module
Table 20-4. DCR n Field Descriptions (continued)
Description
Auto-align. AA and SIZE determine whether the source or destination is auto-aligned, that is, transfers are optimized based on the address and size. See
Section 20.5.4.1, “Auto-Alignment .”
0 Auto-align disabled
1 If SSIZE indicates a transfer no smaller than DSIZE, source accesses are auto-aligned; otherwise, destination accesses are auto-aligned. Source alignment takes precedence over destination alignment. If auto-alignment is enabled, the appropriate address register increments, regardless of DINC or SINC.
Bandwidth control. Indicates the number of bytes in a block transfer. Decoded for internal bandwidth control.
When the byte count reaches a multiple of the BWC value, the DMA releases the bus.the request signal to the internal arbiter, MAP, is negated until the access completes. This lets the arbiter give bus access to another device.
BWC
000
001
010
011
100
101
110
111
Number of kilobytes per block
DMA has priority and does not negate its request until transfer completes.
16 Kbytes
32 Kbytes
64 Kbytes
128 Kbytes
256 Kbytes
512 Kbytes
1024 Kbytes
24-23
22
SINC
21–20
SSIZE
19
DINC
18–17
DSIZE
16
START
Reserved, should be cleared.
Source increment. Controls whether a source address increments after each successful transfer.
0 No change to SAR after a successful transfer.
1 The SAR increments by 1, 2, 4, or 16, as determined by the transfer size.
Source size. Determines the data size of the source bus cycle for the DMA control module.
00 Longword
01 Byte
10 Word
11 Line (16-byte burst)
Destination increment. Controls whether a destination address increments after each successful transfer.
0 No change to the DAR after a successful transfer.
1 The DAR increments by 1, 2, 4, or 16, depending upon the size of the transfer.
Destination size. Determines the data size of the destination bus cycle for the DMA controller.
00 Longword
01 Byte
10 Word
11 Line (16-byte burst)
Start transfer.
0 DMA inactive
1 The DMA begins the transfer in accordance to the values in the control registers. START is cleared automatically after one system clock and is always read as logic 0.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 20-13
DMA Controller Module
Field
15–12
SMOD
Table 20-4. DCR n Field Descriptions (continued)
Description
Source address modulo. Defines the size of the source data circular buffer used by the DMA controller. If enabled
(SMOD is non-zero), the buffer base address will be located on a boundary of the buffer size. The value of this boundary is based upon the initial source address (SAR). The base address should be aligned to a 0-modulo- circular buffer size boundary. Misaligned buffers are not possible. The boundary will be forced to the value determined by the upper address bits in the field selection.
11–8
DMOD
0 Source address modulo feature is disabled.
non-0
The value defines each address bit which is selected to be either the value after next-state SAR calculation is performed or the current SAR register value. This feature provides the ability to easily implement circular data queues. For data queues requiring power-of-2 “size” bytes, the queue should be based at a 0-modulo-size address and the SMOD field set to the appropriate value to effectively freeze the upper address bits. The bit select is defined as ((1 << SMOD+3) - 1) where a resulting 1 in a bit location selects the next state address for the corresponding address bit location and a 0 selects the original register value for the corresponding address bit location.
The modulo implementation includes 32 one-bit 2-to-1 muxes connected to the source address register and the output of an adder used to generate the next-state address. The adder sums the source address register with an offset calculated based on the transfer size and increment control (DCRn[SSIZE, SINC]). The 32-bit wide select vector is formed using the expression noted above and is then used to select either the source address register bit (select = 0) or the corresponding bit position from the adder output (select = 1) for each bit of the next-state source address.
This functionality supports circular queues ranging in size from 16 bytes (SMOD = 1) to 256 KBytes (SMOD = 15).
Destination address modulo. Defines the size of the destination data circular buffer used by the DMA controller.
If enabled (DMOD value is non-zero), the buffer base address will be located on a boundary of the buffer size.
The value of this boundary depends on the initial destination address (DAR). The base address should be aligned to a 0-modulo- circular buffer size boundary. Misaligned buffers are not possible. The boundary will be forced to the value determined by the upper address bits in the field selection.
Same functionality as SMOD, except it is applied to the destination address generation
.
DMOD
0000
0001
0010
...
1111
Circular Buffer Size
Buffer Disabled
16 Bytes
32 Bytes
...
256 Kbytes
20-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
DMA Controller Module
Table 20-4. DCR n Field Descriptions (continued)
Field
7
D_REQ
Description
Disable request.
If this flag is set, the DMA hardware automatically clears the corresponding DCR n [EEXT] bit when the byte count register reaches zero.
6
5–4
LINKCC
0 The channel’s EEXT bit is not affected.
1 The channel’s EEXT bit is cleared when the BCR is exhausted.
Reserved, should be cleared.
Link channel control. Allows DMA channels to have their transfers linked. The current DMA channel will trigger a
DMA request to the linked channels (LCH1 or LCH2) depending on the condition described by the LINKCC bits.
The LINKCC field provides a 2-bit encoded value defining the applicable channel-to-channel linking.
00 No channel-to-channel linking
01 Perform a link to channel LCH1 after each cycle-steal transfer followed by a link to LCH2 after the BCR decrements to zero.
10 Perform a link to channel LCH1 after each cycle-steal transfer
11 Perform a link to channel LCH1 after the BCR decrements to zero
3–2
LCH1
1–0
LCH2
If not in cycle steal mode (DCR n [CS]=0) and LINKCC=01 or 10, then no link to LCH1 will occur.
The link channel (LCH1, LCH2) number cannot be the same as the currently-executing channel, and generates a configuration error if this is attempted (DSRn[CE] is set). This type of configuration error sets both DSRn[7:6].
The DSRn[CE] bit is set the standard configuration error flag, while DSRn[7] is a new flag to explicitly signal an linking configuration error.
If LINKCC = 01, a link to LCH1 is created after each cycle-steal transfer performed by the current DMA channel except the last one is completed. As the last cycle-steal is performed and the BCR reaches zero is done, then the link to LCH1 is closed and a link to LCH2 is created.
If the LINKCC field is non-zero, the contents of the bandwidth control field (DCRn[BWC]) are ignored and effectively forced to zero by the DMA hardware. This is done to prevent any non-zero bandwidth control settings from allowing channel arbitration while any type of link is to be performed.
Link channel 1. Indicates the DMA channel assigned as link channel 1. The link channel number cannot be the same as the currently executing channel, and generates a configuration error if this is attempted (DSR n [CE] is set).
The 2-bit link channel number under control of the encoded link control field, LINKCC.
00 DMA Channel 0
01 DMA Channel 1
10 DMA Channel 2
11 DMA Channel 3
Link channel 2. Indicates the DMA channel assigned as link channel 2. The link channel number cannot be the same as the currently executing channel, and generates a configuration error if this is attempted (DSR n [CE] is set).
The 2-bit link channel number under control of the encoded link control field, LINKCC.
00 DMA Channel 0
01 DMA Channel 1
10 DMA Channel 2
11 DMA Channel 3
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 20-15
DMA Controller Module
20.5
Functional Description
In the following discussion, the term ‘DMA request’ implies that DCR n [START] or DCR n [EEXT] is set, followed by assertion of an internal or external DMA request. The START bit is cleared when the channel begins an internal access.
Before initiating a dual-address access, the DMA module verifies that DCR n [SSIZE,DSIZE] are consistent with the source and destination addresses. If they are not consistent, the configuration error bit,
DSR n [CE], is set. If misalignment is detected, no transfer occurs, DSR n [CE] is set, and, depending on the
DCR configuration, an interrupt event is issued. Note that if the auto-align bit, DCR n [AA], is set, error checking is performed on the appropriate registers.
A read/write transfer reads bytes from the source address and writes them to the destination address. The number of bytes is the larger of the sizes specified by DCR n [SSIZE] and DCR n [DSIZE]. See
Section 20.4.5, “DMA Control Registers (DCR0–DCR3) .”
Source and destination address registers (SAR n and DAR n ) can be programmed in the DCR n to increment at the completion of a successful transfer. BCR n decrements when an address transfer write completes for a single-address access (DCR n [SAA] = 0), or when SAA = 1.A successful address phase implies that
XMAP is asserted, XMAH is negated, and the current data phase is not incurring wait states. In the case of a Version 2 I
2
C, the address phase and data phase are identical;
20.5.1
Transfer Requests (Cycle-Steal and Continuous Modes)
The DMA channel supports internal and external requests. A request is issued by setting DCR n [START] or when a UART or DMA timer asserts a DMA request. Setting DCR n [EEXT] enables recognition of external DMA requests. Selecting between cycle-steal and continuous modes minimizes bus usage for either internal or external requests.
• Cycle-steal mode (DCR n [CS] = 1)—Only one complete transfer from source to destination occurs for each request. If DCR n [EEXT] is set, a request can be either internal or external. An internal request is selected by setting DCR n [START]. An external request is initiated by an on-chip peripheral while DCR n [EEXT] is set.
• Continuous mode (DCR n [CS] = 0)—After an internal or external request, the DMA continuously transfers data until BCR n reaches zero or a multiple of DCR n [BWC] or until DSR n [DONE] is set.
If BCR n is a multiple of BWC, the DMA request signal is negated until the bus cycle terminates to allow the internal arbiter to switch masters. DCR n [BWC] = 000 specifies the maximum transfer rate; other values specify a transfer rate limit.
The DMA performs the specified number of transfers, then relinquishes bus control. The DMA negates its internal bus request on the last transfer before BCR n reaches a multiple of the boundary specified in BWC. Upon completion, the DMA reasserts its bus request to regain mastership at the earliest opportunity. The DMA loses bus control for a minimum of one bus cycle.
20-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
DMA Controller Module
20.5.2
Dual-Address Data Transfer Mode
Each channel supports dual-address transfers. Dual-address transfers consist of a source data read and a destination data write. The DMA controller module begins a dual-address transfer sequence during a DMA request. If no error condition exists, DSR n [REQ] is set.
• Dual-address read—The DMA controller drives the SAR n value onto the internal address bus. If
DCR n [SINC] is set, the SAR n increments by the appropriate number of bytes upon a successful read cycle. When the appropriate number of read cycles complete (multiple reads if the destination size is larger than the source), the DMA initiates the write portion of the transfer.
If a termination error occurs, DSR n [BES,DONE] are set and DMA transactions stop.
• Dual-address write—The DMA controller drives the DAR n value onto the address bus. If
DCR n [DINC] is set, DAR n increments by the appropriate number of bytes at the completion of a successful write cycle. BCR n decrements by the appropriate number of bytes. DSR n [DONE] is set when BCR n reaches zero. If the BCR n is greater than zero, another read/write transfer is initiated.
If the BCR n is a multiple of DCR n [BWC], the DMA request signal is negated until termination of the bus cycle to allow the internal arbiter to switch masters. In the case of a Version 2 I
2
C, there is an idle clock before the next assertion of MTS
If a termination error occurs, DSR n [BED,DONE] are set and DMA transactions stop.
20.5.3
Channel Initialization and Startup
Before a block transfer starts, channel registers must be initialized with information describing configuration, request-generation method, and the data block.
20.5.3.1
Channel Prioritization
The four DMA channels are prioritized in ascending order (channel 0 having highest priority and channel
3 having the lowest) or in an order determined by DCR n [BWC]. If the BWC encoding for a DMA channel is 000, that channel has priority only over the channel immediately preceding it. For example, if
DCR3[BWC] = 000, DMA channel 3 has priority over DMA channel 2 (assuming DCR2[BWC] ≠ 000), but not over DMA channel 1.
If DCR0[BWC] = DCR1[BWC] = 000, DMA0 still has priority over DMA1. In this case, DCR1[BWC] =
000 does not affect prioritization.
Simultaneous external requests are prioritized either in ascending order or in an order determined by each channel’s DCR n [BWC] bits.
20.5.3.2
Programming the DMA Controller Module
Note the following general guidelines for programming the DMA:
• No mechanism exists within the DMA module itself to prevent writes to control registers during
DMA accesses.
• If the DCR n [BWC] value of sequential channels are equal, the channels are prioritized in ascending order.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 20-17
DMA Controller Module
The DMAREQC register is configured to assign peripheral DMA requests to the individual DMA channels.
The SAR n is loaded with the source (read) address. If the transfer is from a peripheral device to memory, the source address is the location of the peripheral data register. If the transfer is from memory to either a peripheral device or memory, the source address is the starting address of the data block. This can be any aligned byte address.
The DAR n should contain the destination (write) address. If the transfer is from a peripheral device to memory, or from memory to memory, the DAR n is loaded with the starting address of the data block to be written. If the transfer is from memory to a peripheral device, DAR n is loaded with the address of the peripheral data register. This address can be any aligned byte address.
SAR n and DAR n change after each cycle depending on DCR n [SSIZE,DSIZE,
SINC,DINC,SMOD,DMOD] and on the starting address. Increment values can be 1, 2, 4, or 16 for byte, word, longword, or 16-byte line transfers, respectively. If the address register is programmed to remain unchanged (no count), the register is not incremented after the data transfer.
BCR n [BCR] must be loaded with the number of byte transfers to occur. It is decremented by 1, 2, 4, or 16 at the end of each transfer, depending on the transfer size. DSR n [DONE] must be cleared for channel startup.
As soon as the channel has been initialized, it is started by writing a one to DCR n [START] or a peripheral
DMA request, depending on the status of DCR n [EEXT]. Programming the channel for internal requests causes the channel to request the bus and start transferring data immediately. If the channel is programmed for external request, a peripheral DMA request must be asserted before the channel requests the bus.
Changes to DCR n are effective immediately while the channel is active. To avoid problems with changing a DMA channel setup, write a one to DSR n [DONE] to stop the DMA channel.
20.5.4
Data Transfer
This section describes auto-alignment and bandwidth control for DMA transfers.
20.5.4.1
Auto-Alignment
Auto-alignment allows block transfers to occur at the optimal size based on the address, byte count, and programmed size. To use this feature, DCR n [AA] must be set. The source is auto-aligned if DCR n [SSIZE] indicates a transfer size larger than DCR n [DSIZE]. Source alignment takes precedence over the destination when the source and destination sizes are equal. Otherwise, the destination is auto-aligned. The address register chosen for alignment increments regardless of the increment value. Configuration error checking is performed on registers not chosen for alignment.
If BCR n is greater than 16, the address determines transfer size. Bytes, words, or longwords are transferred until the address is aligned to the programmed size boundary, at which time accesses begin using the programmed size.
If BCR n is less than 16 at the start of a transfer, the number of bytes remaining dictates transfer size. For example, AA = 1, SAR n = 0x0001, BCR n = 0x00F0, SSIZE = 00 (longword), and DSIZE = 01 (byte).
20-18
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
DMA Controller Module
Because SSIZE > DSIZE, the source is auto-aligned. Error checking is performed on destination registers.
The access sequence is as follows:
1. Read byte from 0x0001—write 1 byte, increment SAR n .
2. Read word from 0x0002—write 2 bytes, increment SAR n .
3. Read longword from 0x0004—write 4 bytes, increment SAR n .
4. Repeat longwords until SAR n = 0x00F0.
5. Read byte from 0x00F0—write byte, increment SAR n .
If DSIZE is another size, data writes are optimized to write the largest size allowed based on the address, but not exceeding the configured size.
20.5.4.2
Bandwidth Control
Bandwidth control makes it possible to force the DMA off the bus to allow access to another device.
DCR n [BWC] provides seven levels of block transfer sizes. If the BCR n decrements to a multiple of the decode of the BWC, the DMA bus request negates until the bus cycle terminates. If a request is pending, the arbiter may then pass bus mastership to another device. If auto-alignment is enabled, DCR n [AA] = 1, the BCR n may skip over the programmed boundary, in which case, the DMA bus request is not negated.
If BWC = 000, the request signal remains asserted until BCR n reaches zero. DMA has priority over the core. Note that in this scheme, the arbiter can always force the DMA to relinquish the bus. See
Section 10.6.3, “Bus Master Park Register (MPARK).”
20.5.5
Termination
An unsuccessful transfer can terminate for one of the following reasons:
•
Error conditions—When the DMA encounters a read or write cycle that terminates with an error
condition, DSR n [BES] is set for a read and DSR n [BED] is set for a write before the transfer is halted. If the error occurred in a write cycle, data in the internal holding register is lost.
• Interrupts—If DCR n [INT] is set, the DMA drives the appropriate internal interrupt signal. The processor can read DSR n to determine whether the transfer terminated successfully or with an error. DSR n [DONE] is then written with a one to clear the interrupt and the DONE and error bits.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 20-19
DMA Controller Module
20-20
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Chapter 21
EzPort
EzPort is a serial Flash programming interface that allows the Flash memory contents on a 32-bit general purpose microcontroller to be read, erased, and programmed from off-chip in a compatible format to many standalone Flash memory chips.
21.1
Features
The EzPort includes the following features:
• Serial interface that is compatible with a subset of the SPI format
• Ability to read, erase, and program Flash memory
• Ability to reset the micro-controller, allowing it to boot from the Flash memory after the memory has been configured
The EzPort allows the Flash memory internal to the controller to be programmed like standard SPI flash memories available from ST Microelectronics, Macronix, Spansion, and other vendors. The EzPort implements the same command set as devices from these vendors, so existing microcontroller or automated test equipment code used to program these devices can also be used to program the device with little or no modification. In essence, the EzPort eliminates the need to use the background debug mode interface to download and run user-developed flash programming code to initialize
21.2
Modes of Operation
The EzPort can operate in one of two different modes:
• Enabled—When enabled, the EzPort steals access to the Flash memory, preventing access from other cores or peripherals. The rest of the micro-controller is disabled when the EzPort is enabled to avoid conflicts.
• Disabled—When the EzPort is disabled, the rest of the micro-controller can access Flash memory as normal.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
21-1
EzPort
is a block diagram of the EzPort.
EzPort Enabled
G
Flash
Controller
EzPort
EZPCS
EZPCK
EZPD
EZPQ
Flash Memory
Microcontroller
Core
Reset
Reset Out
Reset Controller
Figure 21-1. EzPort Block Diagram
21.3
External Signal Description
21.3.1
Overview
contains a list of EzPort external signals.
Name
EZPCK
EZPCS
EZPD
EZPQ
Table 21-1. Signal Descriptions
Description
EzPort Clock
EzPort Chip Select
EzPort Serial Data In
EzPort Serial Data Out
I/O
Input
Input
Input
Output
21.3.2
Detailed Signal Descriptions
21.3.2.1
EZPCK — EzPort Clock
EzPort clock (EZPCK) is the serial clock for data transfers. Serial data in (EZPD) and chip select (EZPCS) are registered on the rising edge of EZPCK while serial data out (EZPQ) is driven on the falling edge of
21-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
EzPort
EZPCK.The maximum frequency of the EzPort clock is half the system clock frequency for all commands except when executing the read data command. When executing the Read Data command, the EzPort clock has a maximum frequency of one eighth the system clock frequency.
21.3.2.2
EZPCS — EzPort Chip Select
EzPort chip select (EZPCS) is the chip select for signalling the start and end of serial transfers. If EZPCS is asserted during and when the micro-controller’s reset out signal is negated, then EzPort is enabled out of reset; otherwise it is disabled. Once EzPort is enabled, asserting EZPCS commences a serial data transfer, which continues until EZPCS is negated again. The negation of EZPCS indicates the current command is finished and resets the EzPort state machine so that it is ready to receive the next command.
21.3.2.3
EZPD — EzPort Serial Data In
EzPort serial data in (EZPD) is the serial data in for data transfers. It is registered on the rising edge of
EZPCK. All commands, addresses, and data are shifted in most significant bit first. When EzPort is driving output data on EZPQ, the data shifted in EZPD is ignored.
21.3.2.4
EZPQ — EzPort Serial Data Out
EzPort serial data out (EZPQ) is the serial data out for data transfers. It is driven on the falling edge of
EZPCK. It is tri-stated, unless EZPCS is asserted and the EzPort is driving data out. All data is shifted out most significant bit first.
21.4
Command Definition
The EzPort receives commands from an external device and translates those commands into Flash memory accesses.
lists the supported commands.
Table 21-2. EzPort Commands
Command Description Code
Address
Bytes
Dummy
Bytes
0 0 WREN
WRDI
RDSR
WRCR
READ
FAST_READ
PP
SE
BE
Write Enable
Write Disable
Read Status Register
Write Config Register
Read Data
Read Data at High Speed
Page Program
Sector Erase
Bulk Erase
0x06
0x04
0x05
0x01
0x03
0x0B
0x02
0xD8
0xC7
0
0
0
3
3
3
3
0
RESET Reset Chip 0xB9 0
1
Lists the compatible commands on the ST Microelectronics Serial Flash Memory parts.
0
0
1
0
0
0
0
0
0
Data
Bytes
0
1
1+
0
0
0
1
1+
4 to 256
0
Compatible
Commands
1
WREN
WRDI
RDSR
WRSR
READ
FAST_READ
PP
SE
BE
DP
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
21-3
EzPort
21.4.1
Command Descriptions
21.4.1.1
Write Enable
The Write Enable command sets the write enable register bit in the status register. The write enable bit must be set for a Write Configuration Register (WRCR), Page Program (PP), Sector Erase (SE), or Bulk Erase
(BE) command to be accepted. The write enable register bit clears on reset, on a Write Disable command, and at the completion of a write, program, or erase command.
This command should not be used if a write is already in progress.
21.4.1.2
Write Disable
The Write Disable command clears the write enable register bit in the status register.
This command should not be used if a write is already in progress.
21.4.1.3
Read Status Register
The Read Status Register command returns the contents of the EzPort Status register.
IPSBAR
Offset:
R
W
Reset:
7
FS
6
WEF
5
CRL
4 3
0/1
1
0 0 0 0
Figure 21-2. EzPort Status Register
1
Reset value reflects if Flash Security is enabled or disabled out of reset.
2
0
Access: read/write
1
WEN
0
0
WIP
0
21-4
Field
7
FS
6
WEF
Table 21-3. EzPort Status Register Field Description
Descriptions
Flash Security. Status flag that indicates if the Flash memory is in secure mode. In secure mode, the following commands are not accepted: Read (READ), Fast Read (FAST_READ), Page Program (PP), Sector Erase
(SE). Secure mode can be exited by performing a Bulk Erase (BE) command, which erases the entire contents of the Flash memory.
0 Flash is not in secure mode.
1 Flash is in secure mode.
Write Error Flag. Status flag that indicates if there has been an error with an erase or program instruction inside the Flash controller due to attempting to program or erase a protected sector, or if there is an error in the Flash memory after performing a Bulk Erase command. The flag clears after a Read Status Register (RDSR) command.
0 No error on previous erase/program command.
1 Error on previous erase/program command.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Field
5
CRL
EzPort
Table 21-3. EzPort Status Register Field Description (continued)
Descriptions
Configuration Register Loaded. Status flag that indicates if the configuration register has been loaded. The configuration register initializes the Flash controllers clock configuration register to generate a divided down clock from the system clock that runs at a frequency of 150 kHz to 200 kHz. This register must be initialized before any erase or program commands are accepted.
0 Configuration register has not been loaded; erase and program commands are not accepted.
1 Configuration register has been loaded; erase and program commands are accepted.
Reserved, should be cleared.
4–2
—
1
WEN
0
WIP
Write Enable. Control bit that must be set before a Write Configuration Register (WRCR), Page Program (PP),
Sector Erase (SE), or Bulk Erase (BE) command is accepted. Is set by the Write Enable (WREN) command and cleared by reset or a Write Disable (WRDI) command. It also clears on completion of a write, erase, or program command.
0 Disables the following write, erase, or program command.
1 Enables the following write, erase, or program command.
Write In Progress. Status flag that sets after a Write Configuration Register (WRCR), Page Program (PP),
Sector Erase (SE), or Bulk Erase (BE) command is accepted and clears once the Flash memory erase or program is completed. Only the Read Status Register (RDSR) command is accepted while a write is in progress.
0 Write is not in progress. Accept any command.
1 Write is in progress. Only accept RDSR command.
21.4.1.4
Write Configuration Register
The Write Configuration Command updates the Flash controller’s clock configuration register. The clock configuration register divides down the Flash controller’s internal system clock to a 150 kHz to 200 kHz clock. This register must be initialized before any erase or program commands are issued to the Flash controller.
This command should not be used if the write error flag is set, a write is in progress, or the configuration register has already been loaded (as it is a write-once register).
IPSBAR
Offset:
Access: read/write
7 6 5 4 3 2 1 0
R
W
Reset: 0
PRDIV8
0 0 0 0
DIV[5:0]
Figure 21-3. EzPort Configuration Register
0 0 0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
21-5
EzPort
Field
7
—
6
PRDIV
5–0
DIV[5:0]
Table 21-4. EzPort Configuration Register Field Description
Reserved, should be cleared.
Descriptions
Enables prescaler divide by 8.
0 The system clock is fed directly into the divider.
1 Enables a prescaler that divides the system clock by 8 before it enters the divider.
Clock divider field. The combination of PRDIV8 and DIV[5:0] effectively divides the system clock down to a frequency between 150 kHz and 200 kHz.
21.4.1.5
Read Data
The Read Data command returns data from the Flash memory, starting at the address specified in the command word. Data will continue being returned for as long as the EzPort chip select (EZPCS) is asserted, with the address automatically incrementing. When the address reaches the highest Flash memory address, it will wrap around to the lowest Flash memory address. In this way, the entire contents of the Flash memory can be returned by one command.
For this command to return the correct data, the EzPort Clock (EZPCK) must run at no more than divide by eight of the internal system clock.
This command should not be used if the write error flag is set, or a write is in progress. This command is not accepted if Flash security is enabled.
21.4.1.6
Read Data at High Speed
This command is identical to the Read Data command, except for the inclusion of a dummy byte following the address bytes and before the first data byte is returned.
This allows the command to run at any frequency of the EzPort Clock (EZPCK) up to and including half the internal system clock frequency of the micro-controller.This command should not be used if the write error flag is set, or a write is in progress. This command is not accepted if Flash security is enabled.
21.4.1.7
Page Program
The Page Program command programs locations in Flash memory that have previously been erased. The starting address of the memory to program is sent after the command word and must be a 32-bit aligned address (the two LSBs must be zero). After every four bytes of data are received by the EzPort, that 32-bit word is programmed into Flash memory with the address automatically incrementing after each write. For this reason, the number of bytes to program must be a multiple of four. Only a maximum of 256 bytes can be programmed at a time; when the address reaches the highest address within any given 256-byte space of memory, it will wrap around to the lowest address in that same space.
This command should not be used if the write error flag is set, a write is in progress, the write enable bit is not set, or the configuration register has not been written.This command is not accepted if Flash security is enabled.
The write error flag will set if there is an attempt to program a protected area of the Flash memory.
21-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
EzPort
21.4.1.8
Sector Erase
The Sector Erase command erases the contents of a 2-Kbyte space of Flash memory. The 3-byte address sent after the command byte can be any address within the space to erase.
This command should not be used if the write error flag is set, a write is in progress, the write enable bit is not set, or the configuration register has not been written. This command is not accepted if Flash security is enabled.
The write error flag will set if there is an attempt to erase a protected area of the Flash memory.
21.4.1.9
Bulk Erase
The Bulk Erase command erases the entire contents of Flash memory, ignoring any protected sectors or
Flash security. The write error flag will set if the Bulk Erase command does not successfully erase the entire contents of Flash memory. Flash security will be disabled if the Bulk Erase command is followed by a Reset Chip command.
This command should not be used if the write error flag is set, a write is in progress, the write enable bit is not set, or the configuration register has not been written.
21.4.1.10 Reset Chip
The Reset Chip command forces the chip into the reset state. If the EzPort chip select (EZPCS) pin is asserted at the end of the reset period, then EzPort will be enabled; otherwise it will be disabled.
This command allows the chip to boot up from Flash memory after it has been programmed by an external source.
This command should not be used if a write is in progress.
21.5
Functional Description
The EzPort provides a simple interface to connect an external device to the Flash memory on board a 32 bit microcontroller.
The interface itself is compatible with the SPI interface (with the EzPort operating as a slave) running in either of the two following modes with data transmitted most significant bit first:
• CPOL = 0, CPHA = 0
• CPOL = 1, CPHA = 1
Commands are issued by the external device to erase, program, or read the contents of the Flash memory.
The serial data out from the EzPort is tri-stated unless data is being driven, allowing the signal to be shared among several different EzPort (or compatible) devices in parallel, provided they have different chip selects.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
21-7
EzPort
21.6
Initialization/Application Information
Prior to issuing any program or erase commands, the clock configuration register must be written to set the Flash state machine clock (FCLK). The Flash controller module runs at the system clock frequency divide by 2, but FCLK must be divided down from this frequency to a frequency between 150 kHz and
200 kHz. Use the following procedure to set the PRDIV8 and DIV[5:0] bits in the clock configuration register.
1. If f
SYS
is greater than 25.6 MHz, PRDIV8 = 1; otherwise PRDIV8 = 0.
2. Determine DIV[5:0] by using the following equation. Keep only the integer portion of the result and discard any fraction. Do not round the result.
DIV =
(
Fsys
2x200kHzx 1 + ( PRDIV8x7 )
3. Thus the Flash state machine clock will be:
Fclk =
) (
Fsys x 1 + PRDIV8x7 ) )
So, for Fsys = 66 MHz, writing 0x54 to the clock configuration register will set Fclk to 196.43 kHz, which is a valid frequency for the timing of program and erase operations.
For proper program and erase operations, it is critical to set Fclk between 150 kHz and 200 kHz. Array damage due to overstress can occur when Fclk is less than 150 kHz. Incomplete programming and erasure can occur when Fclk is greater than 200 kHz.
21-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 22
Programmable Interrupt Timer Modules (PIT0–PIT1)
22.1
Introduction
This chapter describes the operation of the two programmable interrupt timer modules: PIT0–PIT1.
22.1.1
Overview
Each PIT is a 16-bit timer that provides precise interrupts at regular intervals with minimal processor intervention. The timer can either count down from the value written in the modulus register, or it can be a free-running down-counter.
22.1.2
Block Diagram
Internal Bus
Internal Bus
Clock (f sys/
)
Prescaler
EN
PRE[3:0]
DOZE
DBG
16-bit PCNTR n
16-bit PIT Counter
OVW
16-bit PMR n
COUNT = 0
Load
Counter
RLD
PIF
PIE
To Interrupt
Controller
Internal Bus
Figure 22-1. PIT Block Diagram
22.1.3
Low-Power Mode Operation
This subsection describes the operation of the PIT modules in low-power modes and debug mode of operation. Low-power modes are described in the power management module, Chapter 8, “Power
Management.”
shows the PIT module operation in low-power modes, and how it can exit from each mode.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
22-1
Programmable Interrupt Timer Modules (PIT0–PIT1)
NOTE
The low-power interrupt control register (LPICR) in the system control module specifies the interrupt level at or above which the device can be brought out of a low-power mode.
Table 22-1. PIT Module Operation in Low-power Modes
Low-power Mode
Wait
Doze
Stop
Debug
PIT Operation Mode Exit
Normal N/A
Normal if PCSR n [DOZE] cleared, stopped otherwise
Any interrupt at or above level in LPICR, will exit doze mode if PCSR n [DOZE] is set. Otherwise interrupt assertion has no effect.
Stopped No
Normal if PCSR n [DBG] cleared, stopped otherwise
No. Any interrupt will be serviced upon normal exit from debug mode
In wait mode, the PIT module continues to operate as in run mode and can be configured to exit the low-power mode by generating an interrupt request. In doze mode with the PCSR n [DOZE] bit set, PIT module operation stops. In doze mode with the PCSR n [DOZE] bit cleared, doze mode does not affect PIT operation. When doze mode is exited, the PIT continues to operate in the state it was in prior to doze mode.
In stop mode, the internal bus clock is absent, and PIT module operation stops.
In debug mode with the PCSR n [DBG] bit set, PIT module operation stops. In debug mode with the
PCSR n [DBG] bit cleared, debug mode does not affect PIT operation. When debug mode is exited, the PIT continues to operate in its pre-debug mode state, but any updates made in debug mode remain.
22.2
Memory Map/Register Definition
This section contains a memory map, shown in Table 22-2
, and describes the register structure for
PIT0–PIT1.
Table 22-2. Programmable Interrupt Timer Modules Memory Map
IPSBAR Offset
PIT 0
PIT 1
Register
Width
(bits)
Access
Supervisor Access Only Registers
2
PIT Control and Status Register (PCSR n ) 16 R/W 0x15_0000
0x16_0000
0x15_0002
0x16_0002
PIT Modulus Register (PMR n ) 16 R/W
1
Reset Value Section/Page
0x0000
0xFFFF
0x15_0004
0x16_0004
User/Supervisor Access Registers
PIT Count Register (PCNTR n ) 16 R 0xFFFF
22-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Programmable Interrupt Timer Modules (PIT0–PIT1)
1
2
Accesses to reserved address locations have no effect and result in a cycle termination transfer error.
User mode accesses to supervisor only addresses have no effect and result in a cycle termination transfer error.
22.2.1
PIT Control and Status Register (PCSR
n
)
The PCSR n registers configure the corresponding timer’s operation.
IPSBAR
Offset:
0x15_0000 (PCSR0)
0x16_0000 (PCSR
1
)
15
R 0
W
Reset 0
14
0
0
13
0
0
12
0
11 10
PRE
9 8 7
0
6 5 4
0 0 0 0 0 0 0 0 0
Figure 22-2. PIT Control and Status Register (PCSR n )
3
DOZE DBG OVW PIE
0
Access: Supervisor read/write
1 0 2
PIF w1c
0
RLD
0
EN
0
Table 22-3. PCSR n Field Descriptions
Field Description
15–12 Reserved, should be cleared.
11–8
PRE
Prescaler. The read/write prescaler bits select the internal bus clock divisor to generate the PIT clock. To accurately predict the timing of the next count, change the PRE[3:0] bits only when the enable bit (EN) is clear. Changing
PRE[3:0] resets the prescaler counter. System reset and the loading of a new value into the counter also reset the prescaler counter. Setting the EN bit and writing to PRE[3:0] can be done in this same write cycle. Clearing the EN bit stops the prescaler counter.
PRE
0000
0001
0010
...
1101
1110
1111
Internal Bus Clock
Divisor
2
0
2
1
2
2
...
2
13
2
14
2
15
Decimal
Equivalent
1
2
4
...
8192
16384
32768
7
6
DOZE
Reserved, should be cleared.
Doze mode bit. The read/write DOZE bit controls the function of the PIT in doze mode. Reset clears DOZE.
0 PIT function not affected in doze mode
1 PIT function stopped in doze mode. When doze mode is exited, timer operation continues from the state it was in before entering doze mode.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
22-3
Programmable Interrupt Timer Modules (PIT0–PIT1)
Table 22-3. PCSR n Field Descriptions (continued)
Field
5
DBG
4
OVW
3
PIE
2
PIF
1
RLD
0
EN
Description
Debug mode bit. Controls the function of the PIT in halted/debug mode. Reset clears DBG. During debug mode, register read and write accesses function normally. When debug mode is exited, timer operation continues from the state it was in before entering debug mode, but any updates made in debug mode remain.
0 PIT function not affected in debug mode
1 PIT function stopped in debug mode
Note: Changing the DBG bit from 1 to 0 during debug mode starts the PIT timer. Likewise, changing the DBG bit from
0 to 1 during debug mode stops the PIT timer.
Overwrite. Enables writing to PMR n to immediately overwrite the value in the PIT counter.
0 Value in PMR n replaces value in PIT counter when count reaches 0x0000.
1 Writing PMR n immediately replaces value in PIT counter.
PIT interrupt enable. This read/write bit enables the PIF flag to generate interrupt requests.
0 PIF interrupt requests disabled
1 PIF interrupt requests enabled
PIT interrupt flag. This read/write bit is set when the PIT counter reaches 0x0000. Clear PIF by writing a 1 to it or by writing to PMR. Writing 0 has no effect. Reset clears PIF.
0 PIT count has not reached 0x0000.
1 PIT count has reached 0x0000.
Reload bit. The read/write reload bit enables loading the value of PMR n into the PIT counter when the count reaches
0x0000.
0 Counter rolls over to 0xFFFF on count of 0x0000
1 Counter reloaded from PMR n on count of 0x0000
PIT enable bit. Enables PIT operation. When the PIT is disabled, the counter and prescaler are held in a stopped state. This bit is read anytime, write anytime.
0 PIT disabled
1 PIT enabled
22.2.2
PIT Modulus Register (PMR
n
)
The 16-bit read/write PMR n contains the timer modulus value that is loaded into the PIT counter when the count reaches 0x0000 and the PCSR n [RLD] bit is set.
When the PCSR n [OVW] bit is set, PMR n is transparent, and the value written to PMR n is immediately loaded into the PIT counter. The prescaler counter is reset (0xFFFF) anytime a new value is loaded into the PIT counter and also during reset. Reading the PMR n returns the value written in the modulus latch.
Reset initializes PMR n to 0xFFFF.
IPSBAR
Offset:
0x15_0002 (PMR0)
0x16_0002 (PMR 1 )
14 13 15
R
W
Reset 1 1 1
12 11 10 9 8
PM
7 6 5 4
1
3
1 1 1 1 1 1 1 1 1 1 1 1
Figure 22-3. PIT Modulus Register (PMR n )
Access: Supervisor read/write
2 1 0
22-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Programmable Interrupt Timer Modules (PIT0–PIT1)
22.2.3
PIT Count Register (PCNTR
n
)
The 16-bit, read-only PCNTR n contains the counter value. Reading the 16-bit counter with two 8-bit reads is not guaranteed to be coherent. Writing to PCNTR n has no effect, and write cycles are terminated normally.
Access: User read only IPSBAR
Offset:
0x15_0004 (PCNTR0)
0x16_0004 (PCNTR 1 )
14 13 15
R
W
Reset 1 1 1
12 11 10 9 8
PC
7 6 5
1 1 1 1 1 1 1 1
Figure 22-4. PIT Count Register (PCNTR n )
4
1
3
1
2
1
1
1
0
1
22.3
Functional Description
This section describes the PIT functional operation.
22.3.1
Set-and-Forget Timer Operation
This mode of operation is selected when the RLD bit in the PCSR register is set.
When the PIT counter reaches a count of 0x0000, the PIF flag is set in PCSR n . The value in the modulus register is loaded into the counter, and the counter begins decrementing toward 0x0000. If the PCSR n [PIE] bit is set, the PIF flag issues an interrupt request to the CPU.
When the PCSR n [OVW] bit is set, the counter can be directly initialized by writing to PMR n without having to wait for the count to reach 0x0000.
PIT CLOCK
COUNTER
MODULUS
PIF
0x0002 0x0001
0x0005
0x0000 0x0005
Figure 22-5. Counter Reloading from the Modulus Latch
22.3.2
Free-Running Timer Operation
This mode of operation is selected when the PCSR n [RLD] bit is clear. In this mode, the counter rolls over from 0x0000 to 0xFFFF without reloading from the modulus latch and continues to decrement.
When the counter reaches a count of 0x0000, the PCSR n [PIF] flag is set. If the PCSR n [PIE] bit is set, the
PIF flag issues an interrupt request to the CPU.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
22-5
Programmable Interrupt Timer Modules (PIT0–PIT1)
When the PCSR n [OVW] bit is set, the counter can be directly initialized by writing to PMR n without having to wait for the count to reach 0x0000.
PIT CLOCK
COUNTER
MODULUS
PIF
0x0002 0x0001
0x0005
0x0000 0xFFFF
Figure 22-6. Counter in Free-Running Mode
22.3.3
Timeout Specifications
The 16-bit PIT counter and prescaler supports different timeout periods. The prescaler divides the internal bus clock period as selected by the PCSR n [PRE] bits. The PMR n [PM] bits select the timeout period.
22.3.4
Interrupt Operation
shows the interrupt request generated by the PIT.
Table 22-4. PIT Interrupt Requests
Interrupt Request
Timeout
Flag
PIF
Enable Bit
PIE
The PIF flag is set when the PIT counter reaches 0x0000. The PIE bit enables the PIF flag to generate interrupt requests. Clear PIF by writing a 1 to it or by writing to the PMR.
22-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 23
General Purpose Timer Module (GPT)
23.1
Introduction
This device has one 4-channel general purpose timer module (GPT). It consists of a 16-bit counter driven by a 7-stage programmable prescaler.
A timer overflow function allows software to extend the timing capability of the system beyond the 16-bit range of the counter. Each of the four timer channels can be configured for input capture, which can capture the time of a selected transition edge, or for output compare, which can generate output waveforms and timer software delays. These functions allow simultaneous input waveform measurements and output waveform generation.
Additionally, channel 3 can be configured as a 16-bit pulse accumulator that can operate as a simple event counter or as a gated time accumulator. The pulse accumulator uses the GPT channel 3 input/output pin in either event mode or gated time accumulation mode.
23.2
Features
Features of the general-purpose timer include the following:
• Four 16-bit input capture/output compare channels
• 16-bit architecture
• Programmable prescaler
• Pulse widths variable from microseconds to seconds
• Single 16-bit pulse accumulator
• Toggle-on-overflow feature for pulse-width modulator (PWM) generation
• External timer clock input (SYNCA/SYNCB)
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
23-1
General Purpose Timer Module (GPT)
23.3
Block Diagram
CLK[1:0]
System
Clock
Divide by 2
PR[2:0]
Prescaler
PACLK
PACLK/256
PACLK/65536
MUX
X
Channel 3 Output Compare
TCRE
SYNCx
Pin
GPTCNTH:GPTCNTL
16-Bit Counter
Clear Counter
TOF
TOI
CxI
CxF
Interrupt
Logic
Interrupt
Request
TE
Channel 0
16-Bit Comparator
GPTC0H:GPTC0L
16-Bit Latch
CHANNEL 1
16-Bit Comparator
GPTC1H:GPTC1L
16-Bit Latch
C0F
EDG0A
EDG0B
C1F
EDG1A
EDG1B
Edge
Detect
Edge
Detect
IOS0
OM:OL0
TOV0
IOS1
OM:OL1
TOV1
PT0
LOGIC
PT1
LOGIC
CH. 0 Capture
CH. 0 Compare
CH. 1 Capture
CH. 1 Compare
GPTx0
Pin
GPTx1
Pin
Channel 2
Channel3
16-Bit Comparator
GPTC3H:GPTC3L
16-Bit Latch
C3F
EDG3A
EDG3B
Edge
Detect
IOS3
OM:OL3
TOV3
PT3
LOGIC
CH.3 Capture
PA Input
CH. 3 Compare
PAOVF GPTPACNTH:GPTPACNTL
PEDGE
PAE
EDGE
DETECT
PACLK/65536
PACLK/256
Interrupt
Request
PAOVI
PAOVF
16-Bit Counter
Interrupt
Logic
PACLK
MUX
PAMOD
Divide-by-64
PAI
PAIF
Figure 23-1. GPT Block Diagram
PAIF
Divide by 2
GPTx3
Pin
System
Clock
23-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
General Purpose Timer Module (GPT)
23.4
Low-Power Mode Operation
This subsection describes the operation of the general purpose time module in low-power modes and halted mode of operation. Low-power modes are described in Chapter 7, “Power Management
shows the general purpose timer module operation in the low-power modes, and shows how this module may facilitate exit from each mode.
Table 23-1. Watchdog Module Operation in Low-power Modes
Low-power Mode
Wait
Doze
Stop
Halted
Watchdog Operation
Normal
Normal
Stopped
Normal
Mode Exit
No
No
No
No
General purpose timer operation stops in stop mode. When stop mode is exited, the general purpose timer continues to operate in its pre-stop mode state.
23.5
Signal Description
provides an overview of the signal properties.
Table 23-2. Signal Properties
Pin
Name
GPTPORT
Register Bit
Function Reset State Pull-up
GPT0
GPT1
GPT2
PORTT
PORTT
PORTT n n n
0
1
2
GPT channel 0 IC/OC pin
GPT channel 1 IC/OC pin
GPT channel 2 IC/OC pin
Input
Input
Input
Active
Active
Active
GPT3 PORTT n 3 GPT channel 3 IC/OC or PA pin Input Active
SYNC n PORTE[3:0]
1
GPT counter synchronization Input Active
1
SYNCA is available on either PORTE3 or PORTE1; SYNCB is available on either PORTE2 or
PORTE0.
23.5.1
GPT[2:0]
The GPT[2:0] pins are for channel 2–0 input capture and output compare functions. These pins are available for general-purpose input/output (I/O) when not configured for timer functions.
23.5.2
GPT3
The GPT3 pin is for channel 3 input capture and output compare functions or for the pulse accumulator input. This pin is available for general-purpose I/O when not configured for timer functions.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
23-3
General Purpose Timer Module (GPT)
23.5.3
SYNC
n
The SYNC n pin is for synchronization of the timer counter. It can be used to synchronize the counter with externally-timed or clocked events. A high signal on this pin clears the counter.
23.6
Memory Map and Registers
Table 18-4 shows the memory map of the GPT module. The base address for GPT is IPSBAR +
0x1A_0000.
NOTE
Reading reserved or unimplemented locations returns zeros. Writing to reserved or unimplemented locations has no effect.
Table 23-3. QSPI Memory Map
IPSBAR
Offset
1
Register
Width
(bits)
Access Reset Value Section/Page
0x1A_0000 GPT IC/OC Select Register (GPTIOS)
0x1A_0001 GPT Compare Force Register (GPTCFORC)
0x1A_0002 GPT Output Compare 3 Mask Register (GPTOC3M)
0x1A_0003 GPT Output Compare 3 Data Register (GPTOC3D)
0x1A_0004 GPT Counter Register High (GPTCNTH)
2
0x1A_0005 GPT Counter Register Low (GPTCNTL)
2
0x1A_0006 GPT System Control Register 1 (GPTSCR1)
0x1A_0008 GPT Toggle-on-Overflow Register (GPTTOV)
0x1A_0009 GPT Control Register 1 (GPTCTL1)
0x1A_000B GPT Control Register 2 (GPTCTL2)
0x1A_000C GPT Interrupt Enable Register (GPTIE)
0x1A_000D GPT System Control Register 2 (GPTSCR2)
0x1A_000E GPT Flag Register 1 (GPTFLG1)
0x1A_000F GPT Flag Register 2 (GPTFLG2)
0x1A_0010 GPT Channel 0 Register High (GPTC0H)
0x1A_0011 GPT Channel 0 Register Low (GPTC0L)
0x1A_0012 GPT Channel 1 Register High (GPTC1H)
0x1A_0013 GPT Channel 1 Register Low (GPTC1L)
0x1A_0014 GPT Channel 2 Register High (GPTC2H)
0x1A_0015 GPT Channel 2 Register Low (GPTC2L)
0x1A_0016 GPT Channel 3 Register High (GPTC3H)
Supervisor Mode Access Only
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
R/W
R/W
R/W
R/W
R
R
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
23-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
General Purpose Timer Module (GPT)
Table 23-3. QSPI Memory Map (continued)
IPSBAR
Offset
1
Register
0x1A_0017 GPT Channel 3 Register Low (GPTC3L)
Width
(bits)
Access
0x1A_0018 Pulse Accumulator Control Register (GPTPACTL)
0x1A_0019
0x1A_001A Pulse Accumulator Counter Register High
(GPTPACNTH)
2
0x1A_001B Pulse Accumulator Counter Register Low (GPTPACNTL)
2
0x1A_001D GPT Port Data Register (GPTPORT)
0x1A_001E
Pulse Accumulator Flag Register (GPTPAFLG)
GPT Port Data Direction Register (GPTDDR)
8
8
8
8
8
8
8
R/W
R/W
R/W
R/W
R/W
R/W
0x1A_001F GPT Test Register (GPTTST) 8
1
2
Addresses not assigned to a register and undefined register bits are reserved for expansion.
This register is 16 bits wide, and should be read using only word accesses.
Reset Value
0x00
0x00
0x00
0x00
Section/Page
18.6.15/18-19
23.6.1
GPT Input Capture/Output Compare Select Register (GPTIOS)
IPSBAR
Offset:
0x1A_0000 (GPTIOS)
7
0 R
W
Reset:
6
0
5
0
4
0
3 2
Access: Supervisor read/write
IOS
1
0 0 0 0 0 0 0
Figure 23-2. GPT Input Capture/Output Compare Select Register (GPTIOS)
0
0
Table 23-4. GPTIOS Field Descriptions
Field Description
7–4 Reserved, should be cleared.
3–0
IOS
I/O select. The IOS[3:0] bits enable input capture or output compare operation for the corresponding timer channels.
These bits are read anytime (always read 0x00), write anytime.
1 Output compare enabled
0 Input capture enabled
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
23-5
General Purpose Timer Module (GPT)
23.6.2
GPT Compare Force Register (GPCFORC)
IPSBAR
Offset:
0x1A_0001 (GPCFORC)
7
0
6
0 R
W
Reset: 0
5
0
4
0
3 2
Access: Supervisor read/write
FOC
0 0 0 0 0
Figure 23-3. GPT Input Compare Force Register (GPCFORC)
1
0
0
0
Table 23-5. GPTCFORC Field Descriptions
Field Description
7–4 Reserved, should be cleared.
3–0
FOC
Force output compare.Setting an FOC bit causes an immediate output compare on the corresponding channel.
Forcing an output compare does not set the output compare flag. These bits are read anytime, write anytime.
1 Force output compare
0 No effect
NOTE
A successful channel 3 output compare overrides any compare on channels
2:0. For each OC3M bit that is set, the output compare action reflects the corresponding OC3D bit.
23.6.3
GPT Output Compare 3 Mask Register (GPTOC3M)
IPSBAR
Offset:
0x1A_0002 (GPTOC3M)
7
0
6
0 R
W
Reset:
5
0
4
0
3 2
Access: Supervisor read/write
OC3M
1
0 0 0 0 0 0
Figure 23-4. GPT Output Compare 3 Mask Register (GPTOC3M)
0
0
0
23-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
General Purpose Timer Module (GPT)
Table 23-6. GPTOC3M Field Descriptions
Field Description
7–4 Reserved, should be cleared.
3–0
OC3M
Output compare 3 mask. Setting an OC3M bit configures the corresponding PORTT
(IOSx = 1). The OC3M n bits do not change the state of the PORTT n DDR bits. These bits are read anytime, write anytime.
1 Corresponding PORTT n pin configured as output
0 No effect n pin to be an output. OC3M n makes the GPT port pin an output regardless of the data direction bit when the pin is configured for output compare
23.6.4
GPT Output Compare 3 Data Register (GPTOC3D)
IPSBAR
Offset:
0x1A_0003 (GPTOC3D)
7
0
6
0 R
W
Reset:
5
0
4
0
3 2
Access: Supervisor read/write
OC3D
1
0 0 0 0 0 0
Figure 23-5. GPT Output Compare 3 Data Register (GPTOC3D)
0
0
0
Table 23-7. GPTOC3D Field Descriptions
Field Description
7–4 Reserved, should be cleared.
3–0
OC3D
Output compare 3 data. When a successful channel 3 output compare occurs, these bits transfer to the PORTT data register if the corresponding OC3M n bits are set. These bits are read anytime, write anytime.
n
NOTE
A successful channel 3 output compare overrides any channel 2:0 compares.
For each OC3M bit that is set, the output compare action reflects the corresponding OC3D bit.
23.6.5
GPT Counter Register (GPTCNT)
IPSBAR
Offset:
0x1A_0004 (GPTCNT)
14 13 15
R
W
Reset 0 0 0
12 11 10 9 8
CNTR
7 6 5
0 0 0 0 0 0 0 0
Figure 23-6. GPT Counter Register (GPTCNT)
4
0
Access: Supervisor read-only
3
0
2
0
1
0
0
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
23-7
General Purpose Timer Module (GPT)
Table 23-8. GPTCNT Field Descriptions
Field Description
15–0
CNTR
Read-only field that provides the current count of the timer counter. To ensure coherent reading of the timer counter, such that a timer rollover does not occur between two back-to-back 8-bit reads, it is recommended that only word
(16-bit) accesses be used.
A write to GPTCNT may have an extra cycle on the first count because the write is not synchronized with the prescaler clock. The write occurs at least one cycle before the synchronization of the prescaler clock.
These bits are read anytime. They should be written to only in test (special) mode; writing to them has no effect in normal modes.
23.6.6
GPT System Control Register 1 (GPTSCR1)
IPSBAR
Offset:
0x1A_0006 (GPTSCR1)
7 6
0 R
W
Reset:
GPTEN
0
5 4
TFFCA
3
0
2
0
Access: Supervisor read/write
0 0 0 0 0
Figure 23-7. GPT System Control Register 1 (GPTSCR1)
1
0
0
0
0
0
Table 23-9. GPTSCR1 Field Descriptions
Field Description
7
GPTEN
Enables the general purpose timer. When the timer is disabled, only the registers are accessible. Clearing GPTEN reduces power consumption. These bits are read anytime, write anytime.
1 GPT enabled
0 GPT and GPT counter disabled
6–5 Reserved, should be cleared.
4
TFFCA
Timer fast flag clear all. Enables fast clearing of the main timer interrupt flag registers (GPTFLG1 and GPTFLG2) and the PA flag register (GPTPAFLG). TFFCA eliminates the software overhead of a separate clear sequence. See
.
When TFFCA is set:
• An input capture read or a write to an output compare channel clears the corresponding channel flag, CxF.
• Any access of the GPT count registers (GPTCNTH/L) clears the TOF flag.
• Any access of the PA counter registers (GPTPACNT) clears both the PAOVF and PAIF flags in GPTPAFLG.
Writing logic 1s to the flags clears them only when TFFCA is clear.
1 Fast flag clearing
0 Normal flag clearing
3–0 Reserved, should be cleared.
23-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
General Purpose Timer Module (GPT)
Write GPTFLG1 Register
Data Bit n
C n F
Clear
C n F Flag
TFFCA
Read GPTC n Registers
Write GPTC n Registers
Figure 23-8. Fast Clear Flag Logic
23.6.7
GPT Toggle-On-Overflow Register (GPTTOV)
IPSBAR
Offset:
0x1A_0008 (GPTTOV)
7
0 R
W
Reset: 0
6
0
5
0
4
0
3 2
Access: Supervisor read/write
TOV
0 0 0 0 0
Figure 23-9. GPT Toggle-On-Overflow Register (GPTTOV)
1
0
0
0
Table 23-10. GPTTOV Field Description
Field Description
7–4 Reserved, should be cleared.
3–0
TOV
Toggles the output compare pin on overflow for each channel. This feature only takes effect when in output compare mode. When set, it takes precedence over forced output compare but not channel 3 override events. These bits are read anytime, write anytime.
1 Toggle output compare pin on overflow feature enabled
0 Toggle output compare pin on overflow feature disabled
23.6.8
GPT Control Register 1 (GPTCTL1)
IPSBAR
Offset:
0x1A_0009 (GPTCTL1)
7 6
R
W
Reset:
OM3
0
OL3
5
OM2
4
OL2
3
OM1
2
OL1
0 0 0 0 0
Figure 23-10. GPT Control Register 1 (GPTCTL1)
Access: Supervisor read/write
1
OM0
0
0
OL0
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
23-9
General Purpose Timer Module (GPT)
Table 23-11. GPTCL1 Field Descriptions
Field Description
7–0
OMx/OLx
Output mode/output level. Selects the output action to be taken as a result of a successful output compare on each channel. When either OM n or OL n is set and the IOS n bit is set, the pin is an output regardless of the state of the corresponding DDR bit. These bits are read anytime, write anytime.
00 GPT disconnected from output pin logic
01 Toggle OC n output line
10 Clear OC n output line
11 Set OC n line
Note: Channel 3 shares a pin with the pulse accumulator input pin. To use the PAI input, clear both the OM3 and
OL3 bits and clear the OC3M3 bit in the output compare 3 mask register.
23.6.9
GPT Control Register 2 (GPTCTL2)
IPSBAR
Offset:
0x1A_000B (GPTCTL2)
7 6
R
W
Reset:
EDG3B
0
EDG3A
5
EDG2B
4
EDG2A
3
EDG1B
2
Access: Supervisor read/write
EDG1A
0 0 0 0 0
Figure 23-11. GPT Control Register 2(GPTCTL2)
1
EDG0B
0
0
EDG0A
0
Table 23-12. GPTLCTL2 Field Descriptions
Field Description
7–0
EDGn[B:A]
Input capture edge control. Configures the input capture edge detector circuits for each channel. These bits are read anytime, write anytime.
00 Input capture disabled
01 Input capture on rising edges only
10 Input capture on falling edges only
11 Input capture on any edge (rising or falling)
23.6.10 GPT Interrupt Enable Register (GPTIE)
IPSBAR
Offset:
0x1A_000C (GPTIE)
7
0 R
W
Reset: 0
6
0
5
0
4
0
3 2
Access: Supervisor read/write
0 0 0 0 0
Figure 23-12. GPT Interrupt Enable Register (GPTIE)
CI
1
0
0
0
23-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
General Purpose Timer Module (GPT)
Table 23-13. GPTIE Field Descriptions
Field Description
7–4 Reserved, should be cleared.
3–0
C n l
Channel interrupt enable. Enables the C[3:0]F flags in GPT flag register 1 to generate interrupt requests for each channel. These bits are read anytime, write anytime.
1 Corresponding channel interrupt requests enabled
0 Corresponding channel interrupt requests disabled
23.6.11 GPT System Control Register 2 (GPTSCR2)
IPSBAR
Offset:
0x1A_000D (GPTSCR2)
7 6
0 R
W
Reset:
TOI
0
5
PUPT
4
RDPT
3
TCRE
2
Access: Supervisor read/write
0 0 0 0 0
Figure 23-13. GPT System Control Register 2 (GPTSCR2)
1
PR
0
0
0
Table 23-14. GPTSCR2 Field Descriptions
Field
7
TOI
6
5
PUPT
4
RDPT
3
TCRE
Description
Enables timer overflow interrupt requests.
1 Overflow interrupt requests enabled
0 Overflow interrupt requests disabled
Reserved, should be cleared.
Enables pull-up resistors on the GPT ports when the ports are configured as inputs.
1 Pull-up resistors enabled
0 Pull-up resistors disabled
GPT drive reduction. Reduces the output driver size.
1 Output drive reduction enabled
0 Output drive reduction disabled
Enables a counter reset after a channel 3 compare.
1 Counter reset enabled
0 Counter reset disabled
Note: When the GPT channel 3 registers contain 0x0000 and TCRE is set, the GPT counter registers remain at
0x0000 all the time. When the GPT channel 3 registers contain 0xFFFF and TCRE is set, TOF does not get set even though the GPT counter registers go from 0xFFFF to 0x0000.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
23-11
General Purpose Timer Module (GPT)
Field
2–0
PR
Table 23-14. GPTSCR2 Field Descriptions (continued)
Description
Prescaler bits. Select the prescaler divisor for the GPT counter.
000 Prescaler divisor 1
001 Prescaler divisor 2
010 Prescaler divisor 4
011 Prescaler divisor 8
100 Prescaler divisor 16
101 Prescaler divisor 32
110 Prescaler divisor 64
111 Prescaler divisor 128
Note: The newly selected prescaled clock does not take effect until the next synchronized edge of the prescaled clock when the clock count transitions to 0x0000.)
23.6.12 GPT Flag Register 1 (GPTFLG1)
IPSBAR
Offset:
0x1A_000E (GPTFLG1)
7
0
6
0 R
W
Reset: 0
5
0
4
0
3 2
0 0 0 0
Figure 23-14. GPT Flag Register 1 (GPTFLG1)
0
Access: Supervisor read/write
CF
1
0
0
0
Table 23-15. GPTFLG1 Field Descriptions
Field Description
7–4 Reserved, should be cleared.
3–0
C n F
Channel flags. A channel flag is set when an input capture or output compare event occurs. These bits are read anytime, write anytime (writing 1 clears the flag, writing 0 has no effect).
Note: When the fast flag clear all bit, GPTSCR1[TFFCA], is set, an input capture read or an output compare write clears the corresponding channel flag. When a channel flag is set, it does not inhibit subsequent output compares or input captures.
23.6.13 GPT Flag Register 2 (GPTFLG2)
IPSBAR
Offset:
0x1A_000F (GPTFLG2)
7 6
0 R
W
Reset:
TOF
0
5
0
4
0
3 2
0 0 0 0
Figure 23-15. GPT Flag Register 2 (GPTFLG2)
0
Access: Supervisor read/write
CF
1
0
0
0
23-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
General Purpose Timer Module (GPT)
Table 23-16. GPTFLG2 Field Descriptions
Field Description
7
TOF
Timer overflow flag. Set when the GPT counter rolls over from 0xFFFF to 0x0000. If the TOI bit in GPTSCR2 is also set, TOF generates an interrupt request. This bit is read anytime, write anytime (writing 1 clears the flag, and writing
0 has no effect).
1 Timer overflow
0 No timer overflow
Note: When the GPT channel 3 registers contain 0xFFFF and TCRE is set, TOF does not get set even though the
GPT counter registers go from 0xFFFF to 0x0000. When TOF is set, it does not inhibit subsequent overflow events.
6–4 Reserved, should be cleared.
3–0
C n F
Channel flags. A channel flag is set when an input capture or output compare event occurs. These bits are read anytime, write anytime (writing 1 clears the flag, writing 0 has no effect).
Note: When the fast flag clear all bit, GPTSCR1[TFFCA], is set, any access to the GPT counter registers clears GPT flag register 2.
23.6.14 GPT Channel Registers (GPTC
n
)
IPSBAR
Offset:
0x1A_0010, 0x1A_0012, 0x1A_0014, 0x1A_0016, (GPTC n )
14 13 12 11 10 9 8 15
R
W
Reset 0 0 0 0 0 0 0 0
CCNT
7
0
6
0
5
0
Figure 23-16. GPT Channel[0:3] Register (GPTC n )
4
0
3
0
Access: Supervisor read/write
2 1 0
0 0 0
Table 23-17. GPTC n Field Descriptions
Field Description
15–0
CCNT
When a channel is configured for input capture (IOS n = 0), the GPT channel registers latch the value of the free-running counter when a defined transition occurs on the corresponding input capture pin.
When a channel is configured for output compare (IOS n = 1), the GPT channel registers contain the output compare value.
To ensure coherent reading of the GPT counter, such that a timer rollover does not occur between back-to-back 8-bit reads, it is recommended that only word (16-bit) accesses be used. These bits are read anytime, write anytime (for the output compare channel); writing to the input capture channel has no effect.
23.6.15 Pulse Accumulator Control Register (GPTPACTL)
IPSBAR
Offset:
0x1A_0018 (GPTPACTL)
6 7
0 R
W
Reset:
PAE
5
PAMOD
4
PEDGE
3
CLK
2
Access: Supervisor read/write
1
PAOVI
0 0 0 0 0 0
Figure 23-17. Pulse Accumulator Control Register (GPTPACTL)
0
0
PAI
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
23-13
General Purpose Timer Module (GPT)
Table 23-18. GPTPACTL Field Descriptions
Field Description
7
6
PAE
5
PAMOD
Pulse accumulator mode. Selects event counter mode or gated time accumulation mode.
1 Gated time accumulation mode
0 Event counter mode
4
PEDGE
Pulse accumulator edge. Selects falling or rising edges on the PAI pin to increment the counter.
In event counter mode (PAMOD = 0):
1 Rising PAI edge increments counter
0 Falling PAI edge increments counter
In gated time accumulation mode (PAMOD = 1):
1 Low PAI input enables divide-by-64 clock to pulse accumulator and trailing rising edge on PAI sets PAIF flag.
0 High PAI input enables divide-by-64 clock to pulse accumulator and trailing falling edge on PAI sets PAIF flag.
Note: The timer prescaler generates the divide-by-64 clock. If the timer is not active, there is no divide-by-64 clock.
To operate in gated time accumulation mode:
1. Apply logic 0 to RSTI pin.
2. Initialize registers for pulse accumulator mode test.
3. Apply appropriate level to PAI pin.
4. Enable GPT.
3–2
CLK
Select the GPT counter input clock. Changing the CLK bits causes an immediate change in the GPT counter clock input.
00 GPT prescaler clock (When PAE = 0, the GPT prescaler clock is always the GPT counter clock.)
01 PACLK
10 PACLK/256
11 PACLK/65536
1
PAOVI
0
PAI
Reserved, should be cleared.
Enables the pulse accumulator.
1 Pulse accumulator enabled
0 Pulse accumulator disabled
Note: The pulse accumulator can operate in event mode even when the GPT enable bit, GPTEN, is clear.
Pulse accumulator overflow interrupt enable. Enables the PAOVF flag to generate interrupt requests.
1 PAOVF interrupt requests enabled
0 PAOVF interrupt requests disabled
Pulse accumulator input interrupt enable. Enables the PAIF flag to generate interrupt requests.
1 PAIF interrupt requests enabled
0 PAIF interrupt requests disabled
23.6.16 Pulse Accumulator Flag Register (GPTPAFLG)
IPSBAR
Offset:
0x1A_0019 (GPTPAFLG)
7
0
6
0 R
W
Reset: 0
5
0
4
0
3
0
2
0
Access: Supervisor read/write
0 0 0 0 0
Figure 23-18. Pulse Accumulator Flag Register (GPTPAFLG)
1
PAOVF
0
0
PAIF
0
23-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
General Purpose Timer Module (GPT)
Table 23-19. GPTPAFLG Field Descriptions
Field Description
7–2 Reserved, should be cleared.
1
PAOVF
Pulse accumulator overflow flag. Set when the 16-bit pulse accumulator rolls over from 0xFFFF to 0x0000. If the
GPTPACTL[PAOVI] bit is also set, PAOVF generates an interrupt request. Clear PAOVF by writing a 1 to it. This bit is read anytime, write anytime. (Writing 1 clears the flag; writing 0 has no effect.)
1 Pulse accumulator overflow
0 No pulse accumulator overflow
0
PAIF
Pulse accumulator input flag. Set when the selected edge is detected at the PAI pin. In event counter mode, the event edge sets PAIF. In gated time accumulation mode, the trailing edge of the gate signal at the PAI pin sets PAIF. If the
PAI bit in GPTPACTL is also set, PAIF generates an interrupt request. Clear PAIF by writing a 1 to it.
1 Active PAI input
0 No active PAI input
NOTE
When the fast flag clear all enable bit (GPTSCR1[TFFCA]) is set, any access to the pulse accumulator counter registers clears all the flags in
GPTPAFLG.
23.6.17 Pulse Accumulator Counter Register (GPTPACNT)
IPSBAR
Offset:
0x1A_001A (GPTPACNT)
14 13 12 15
R
W
Reset 0 0 0 0
11
0
10
0
9
0
8
0
7
PACNT
0
6
0
5
0
4
0
3
0
Figure 23-19. Pulse Accumulator Counter Register (GPTPACNT)
Access: Supervisor read/write
2 1 0
0 0 0
Table 23-20. GPTPACR Field Descriptions
Field Description
15–0
PACNT
Contains the number of active input edges on the PAI pin since the last reset.
Note: Reading the pulse accumulator counter registers immediately after an active edge on the PAI pin may miss the last count since the input first has to be synchronized with the bus clock.
To ensure coherent reading of the PA counter, such that the counter does not increment between back-to-back 8-bit reads, it is recommended that only word (16-bit) accesses be used. These bits are read anytime, write anytime.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
23-15
General Purpose Timer Module (GPT)
23.6.18 GPT Port Data Register (GPTPORT)
IPSBAR
Offset:
0x1A_001D (GPTPORT)
7
0
6
0 R
W
Reset: 0
5
0
4
0
3 2
0 0 0 0 0
Figure 23-20. GPT Port Data Register (GPTPORT)
Access: Supervisor read/write
PORTT
1
0
0
0
Table 23-21. GPTPORT Field Descriptions
Field Description
7–4 Reserved, should be cleared.
3–0
PORTT
GPT port input capture/output compare data. Data written to GPTPORT is buffered and drives the pins only when they are configured as general-purpose outputs.
Reading an input (DDR bit = 0) reads the pin state; reading an output (DDR bit = 1) reads the latched value. Writing to a pin configured as a GPT output does not change the pin state. These bits are read anytime (read pin state when corresponding PORTT n bit is 0, read pin driver state when corresponding GPTDDR bit is 1), write anytime.
23.6.19 GPT Port Data Direction Register (GPTDDR)
Field
GPT Function
7 6
—
—
5 4 3
Pulse Accumulator Function —
PAI
Reset
R/W
Address
0000_0000
R/W
IPSBAR + 0x1A_001E
Figure 23-21. GPT Port Data Direction Register (GPTDDR)
DDRT
IC/OC
—
Table 23-22. GPTDDR Field Descriptions
Bit(s)
7–4
3–0
Name
—
DDRT
0
Description
Reserved, should be cleared.
Control the port logic of PORTT n . Reset clears the PORTT n data direction register, configuring all GPT port pins as inputs. These bits are read anytime, write anytime.
1 Corresponding pin configured as output
0 Corresponding pin configured as input
23-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
General Purpose Timer Module (GPT)
23.7
Functional Description
The general purpose timer (GPT) module is a 16-bit, 4-channel timer with input capture and output compare functions and a pulse accumulator.
23.7.1
Prescaler
The prescaler divides the module clock by 1 or 16. The PR[2:0] bits in GPTSCR2 select the prescaler divisor.
23.7.2
Input Capture
Clearing an I/O select bit (IOS n ) configures channel n as an input capture channel. The input capture function captures the time at which an external event occurs. When an active edge occurs on the pin of an input capture channel, the timer transfers the value in the GPT counter into the GPT channel registers
(GPTC n ).
The minimum pulse width for the input capture input is greater than two module clocks.
The input capture function does not force data direction. The GPT port data direction register controls the data direction of an input capture pin. Pin conditions such as rising or falling edges can trigger an input capture only on a pin configured as an input.
An input capture on channel n sets the C n F flag. The C n I bit enables the C n F flag to generate interrupt requests.
23.7.3
Output Compare
Setting an I/O select bit (IOS n ) configures channel n as an output compare channel. The output compare function can generate a periodic pulse with a programmable polarity, duration, and frequency. When the
GPT counter reaches the value in the channel registers of an output compare channel, the timer can set, clear, or toggle the channel pin. An output compare on channel n sets the C n F flag. The C n I bit enables the
C n F flag to generate interrupt requests.
The output mode (OM n) and level bits (OL n) select, set, clear, or toggle on output compare. Clearing both
OM n and OL n disconnects the pin from the output logic.
Setting a force output compare bit (FOC n ) causes an output compare on channel n . A forced output compare does not set the channel flag.
A successful output compare on channel 3 overrides output compares on all other output compare channels. A channel 3 output compare can cause bits in the output compare 3 data register to transfer to the GPT port data register, depending on the output compare 3 mask register. The output compare 3 mask register masks the bits in the output compare 3 data register. The GPT counter reset enable bit, TCRE, enables channel 3 output compares to reset the GPT counter. A channel 3 output compare can reset the
GPT counter even if the OC3/PAI pin is being used as the pulse accumulator input.
An output compare overrides the data direction bit of the output compare pin but does not change the state of the data direction bit.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
23-17
General Purpose Timer Module (GPT)
Writing to the PORTT n bit of an output compare pin does not affect the pin state. The value written is stored in an internal latch. When the pin becomes available for general-purpose output, the last value written to the bit appears at the pin.
23.7.4
Pulse Accumulator
The pulse accumulator (PA) is a 16-bit counter that can operate in two modes:
1. Event counter mode: counts edges of selected polarity on the pulse accumulator input pin, PAI
2. Gated time accumulation mode: counts pulses from a divide-by-64 clock
The PA mode bit (PAMOD) selects the mode of operation.
The minimum pulse width for the PAI input is greater than two module clocks.
23.7.5
Event Counter Mode
Clearing the PAMOD bit configures the PA for event counter operation. An active edge on the PAI pin increments the PA. The PA edge bit (PEDGE) selects falling edges or rising edges to increment the PA.
An active edge on the PAI pin sets the PA input flag (PAIF). The PA input interrupt enable bit (PAI) enables the PAIF flag to generate interrupt requests.
NOTE
The PAI input and GPT channel 3 use the same pin. To use the PAI input, disconnect it from the output logic by clearing the channel 3 output mode and output level bits, OM3 and OL3. Also clear the channel 3 output compare 3 mask bit (OC3M3).
The PA counter register (GPTPACNT) reflects the number of active input edges on the PAI pin since the last reset.
The PA overflow flag (PAOVF) is set when the PA rolls over from 0xFFFF to 0x0000. The PA overflow interrupt enable bit (PAOVI) enables the PAOVF flag to generate interrupt requests.
NOTE
The PA can operate in event counter mode even when the GPT enable bit
(GPTEN) is clear.
23.7.6
Gated Time Accumulation Mode
Setting the PAMOD bit configures the PA for gated time accumulation operation. An active level on the
PAI pin enables a divide-by-64 clock to drive the PA. The PA edge bit (PEDGE) selects low levels or high levels to enable the divide-by-64 clock.
The trailing edge of the active level at the PAI pin sets the PA input flag (PAIF). The PA input interrupt enable bit (PAI) enables the PAIF flag to generate interrupt requests.
23-18
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
General Purpose Timer Module (GPT)
NOTE
The PAI input and GPT channel 3 use the same pin. To use the PAI input, disconnect it from the output logic by clearing the channel 3 output mode
(OM3) and output level (OL3) bits. Also clear the channel 3 output compare mask bit (OC3M3).
The PA counter register (GPTPACNT) reflects the number of pulses from the divide-by-64 clock since the last reset.
NOTE
The GPT prescaler generates the divide-by-64 clock. If the timer is not active, there is no divide-by-64 clock.
PULSE
ACCUMULATOR
PAD
CHANNEL 3 OUTPUT COMPARE
OM3
OL3
OC3M3
Figure 23-22. Channel 3 Output Compare/Pulse Accumulator Logic
23.7.7
General-Purpose I/O Ports
An I/O pin used by the timer defaults to general-purpose I/O unless an internal function that uses that pin is enabled.
The PORTT n pins can be configured for either an input capture function or an output compare function.
The IOS n bits in the GPT IC/OC select register configure the PORTT n pins as either input capture or output compare pins.
The PORTT n data direction register controls the data direction of an input capture pin. External pin conditions trigger input captures on input capture pins configured as inputs.
To configure a pin for input capture:
1. Clear the pin’s IOS bit in GPTIOS.
2. Clear the pin’s DDR bit in PORTT n DDR.
3. Write to GPTCTL2 to select the input edge to detect.
PORTT n DDR does not affect the data direction of an output compare pin. The output compare function overrides the data direction register but does not affect the state of the data direction register.
To configure a pin for output compare:
1. Set the pin’s IOS bit in GPTIOS.
2. Write the output compare value to GPTC n .
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
23-19
General Purpose Timer Module (GPT)
3. Clear the pin’s DDR bit in PORTT n DDR.
4. Write to the OM n /OL n bits in GPTCTL1 to select the output action.
shows how various timer settings affect pin functionality.
Table 23-23. GPT Settings and Pin Functions
GPTEN DDR
1
GPTIOS
EDGx
[B:A]
OMx/
OLx
2
OC3Mx
3
Pin
Data
Dir.
Pin
Driven by
Pin
Function
Comments
0 0 X
4
X X X In Ext.
Digital input GPT disabled by GPTEN = 0
0
1
1
1
1
0
1
0
X
0 (IC)
0
0
X
0 (IC disabled)
0
<> 0
X
X
X
X
X
0
0
0
Out Data reg.
Digital output GPT disabled by GPTEN = 0
In Ext.
Digital input Input capture disabled by EDG setting n
Out Data reg.
Digital output Input capture disabled by EDG n setting
In Ext.
IC and digital input
Normal settings for input capture
1
1
1
0
0
0
<> 0
<> 0
X
X
0
1
Out Data reg.
Digital output Input capture of data driven to output pin by CPU
In Ext.
IC and digital input
OC3M setting has no effect because
IOS = 0
1 1 0 <> 0 X 1 Out Data reg.
Digital output OC3M setting has no effect because
IOS = 0; input capture of data driven to output pin by CPU
1
1
1
1
1
0
1
0
1
0
1 (OC)
1
1
1
1
X
(3)
X
X
X
X
0
0
5
<> 0
<> 0
X
0
0
0
0
1
In Ext.
Digital input Output compare takes place but does not affect the pin because of the OM n /OL n setting
Out Data reg.
Digital output Output compare takes place but does not affect the pin because of the OM n /OL n setting
Out OC action Output compare
Pin readable only if DDR = 0
(5)
Out OC action Output compare
Pin driven by OC action
(5)
Out OC action/
OC3D n
Output compare
(ch 3)
Pin readable only if DDR = 0
6
1 1 1 X X 1 Out OC action/
OC3D n
Output compare/
OC3D n
(ch 3)
Pin driven by channel OC action and
OC3D n via channel 3 OC
(6)
1
2
When DDR set the pin as input (0), reading the data register will return the state of the pin. When DDR set the pin as output
(1), reading the data register will return the content of the data latch. Pin conditions such as rising or falling edges can trigger an input capture on a pin configured as an input.
OM n /OL n bit pairs select the output action to be taken as a result of a successful output compare. When either OM n or OL n is set and the IOS n bit is set, the pin is an output regardless of the state of the corresponding DDR bit.
23-20
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
General Purpose Timer Module (GPT)
3
4
5
6
Setting an OC3M bit configures the corresponding PORTT n pin to be output. OC3M n makes the PORTT n pin an output regardless of the data direction bit when the pin is configured for output compare (IOS n = 1). The OC3M n bits do not change the state of the PORTT n DDR bits.
X = Don’t care
An output compare overrides the data direction bit of the output compare pin but does not change the state of the data direction bit. Enabling output compare disables data register drive of the pin.
A successful output compare on channel 3 causes an output value determined by OC3D n value to temporarily override the output compare pin state of any other output compare channel.The next OC action for the specific channel will still be output to the pin. A channel 3 output compare can cause bits in the output compare 3 data register to transfer to the GPT port data register, depending on the output compare 3 mask register.
23.8
Reset
.”
23.9
Interrupts
lists the interrupt requests generated by the timer.
Table 23-24. GPT Interrupt Requests
Interrupt Request
Channel 3 IC/OC
Channel 2 IC/OC
Channel 1 IC/OC
Channel 0 IC/OC
PA overflow
PA input
Timer overflow
Flag
C3F
C2F
C1F
C0F
PAOVF
PAIF
TOF
Enable Bit
C3I
C2I
C1I
C0I
PAOVI
PAI
TOI
23.9.1
GPT Channel Interrupts (C
n
F)
A channel flag is set when an input capture or output compare event occurs. Clear a channel flag by writing a 1 to it.
NOTE
When the fast flag clear all bit (GPTSCR1[TFFCA]) is set, an input capture read or an output compare write clears the corresponding channel flag.
When a channel flag is set, it does not inhibit subsequent output compares or input captures
23.9.2
Pulse Accumulator Overflow (PAOVF)
PAOVF is set when the 16-bit pulse accumulator rolls over from 0xFFFF to 0x0000. If the PAOVI bit in
GPTPACTL is also set, PAOVF generates an interrupt request. Clear PAOVF by writing a 1 to this flag.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
23-21
General Purpose Timer Module (GPT)
NOTE
When the fast flag clear all enable bit (GPTSCR1[TFFCA]) is set, any access to the pulse accumulator counter registers clears all the flags in
GPTPAFLG.
23.9.3
Pulse Accumulator Input (PAIF)
PAIF is set when the selected edge is detected at the PAI pin. In event counter mode, the event edge sets
PAIF. In gated time accumulation mode, the trailing edge of the gate signal at the PAI pin sets PAIF. If the
PAI bit in GPTPACTL is also set, PAIF generates an interrupt request. Clear PAIF by writing a 1 to this flag.
NOTE
When the fast flag clear all enable bit (GPTSCR1[TFFCA]) is set, any access to the pulse accumulator counter registers clears all the flags in
GPTPAFLG.
23.9.4
Timer Overflow (TOF)
TOF is set when the GPT counter rolls over from 0xFFFF to 0x0000. If the GPTSCR2[TOI] bit is also set,
TOF generates an interrupt request. Clear TOF by writing a 1 to this flag.
NOTE
When the GPT channel 3 registers contain 0xFFFF and TCRE is set, TOF does not get set even though the GPT counter registers go from 0xFFFF to
0x0000.
When the fast flag clear all bit (GPTSCR1[TFFCA]) is set, any access to the
GPT counter registers clears GPT flag register 2.
When TOF is set, it does not inhibit future overflow events.
23-22
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 24
DMA Timers (DTIM0–DTIM3)
24.1
Introduction
This chapter describes the configuration and operation of the four direct memory access (DMA) timer modules (DTIM0, DTIM1, DTIM2, and DTIM3). These 32-bit timers provide input capture and reference compare capabilities with optional signaling of events using interrupts or DMA triggers. Additionally, programming examples are included.
NOTE
The designation ‘ n ’ is used throughout this section to refer to registers or signals associated with one of the four identical timer modules: DTIM0,
DTIM1, DTIM2, or DTIM3.
24.1.1
Overview
Each DMA timer module has a separate register set for configuration and control. The timers can be configured to operate from the internal bus clock or from an external clocking source using the DT n IN signal. If the internal bus clock is selected, it can be divided by 16 or 1. The selected clock source is routed to an 8-bit programmable prescaler that clocks the actual DMA timer counter register (DTCN n ). Using the
DTMR n , DTXMR n , DTCR n , and DTRR n registers, the DMA timer may be configured to assert an output signal, generate an interrupt, or initiate a DMA transfer on a particular event.
NOTE
The GPIO module must be configured to enable the peripheral function of the appropriate pins (refer to Chapter 13, “General Purpose I/O Module” ) prior to configuring the DMA Timers.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
24-1
DMA Timers (DTIM0–DTIM3)
is a block diagram of one of the four identical timer modules.
Internal Bus Clock
(÷1 or ÷16 )
DT n IN
DMA Timer
Clock
Generator clock
15 0
DMA Timer Mode Register (DTMR n )
Prescaler Mode Bits
7 0
DMA Timer Extended Mode
Register (DTXMR n )
Divider
Capture
Detection
31
DMA Timer Counter Register (DTCN n )
(contains incrementing value)
0
31
DMA Timer Capture Register (DTCR n )
(latches DTCN value when triggered by DT n IN)
0 31
DMA Timer Reference Register (DTRR n )
(reference value for comparison with DTCN)
0
7
DMA Timer Event Register (DTER n )
(indicates capture or when DTCN = DTRR n )
0
DT n OUT
To Interrupt controller
DMA Request
Figure 24-1. DMA Timer Block Diagram
24.1.2
Features
Each DMA timer module has the following features:
• Maximum timeout period of 266,521 seconds at 66 MHz (~74 hours)
• 15-ns resolution at 66 MHz
• Programmable sources for the clock input, including external clock
• Programmable prescaler
• Input-capture capability with programmable trigger edge on input pin
• Programmable mode for the output pin on reference compare
• Free run and restart modes
• Programmable interrupt or DMA request on input capture or reference-compare
24.2
Memory Map/Register Definition
The timer module registers, shown in Table 24-1
, can be modified at any time.
24-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
DMA Timers (DTIM0–DTIM3)
Table 24-1. DMA Timer Module Memory Map
IPSBAR Offset
DMA Timer 0
DMA Timer 1
DMA Timer 2
DMA Timer 3
0x00_0400
0x00_0440
0x00_0480
0x00_04C0
0x00_0402
0x00_0442
0x00_0482
0x00_04C2
0x00_0403
0x00_0443
0x00_0483
0x00_04C3
0x00_0404
0x00_0444
0x00_0484
0x00_04C4
0x00_0408
0x00_0448
0x00_0488
0x00_04C8
0x00_040C
0x00_044C
0x00_048C
0x00_04CC
Register
DMA Timer n Mode Register (DTMR n )
Width
(bits)
16
DMA Timer n Extended Mode Register (DTXMR n ) 8
DMA Timer n Event Register (DTER n )
DMA Timer n Reference Register (DTRR n )
DMA Timer n Capture Register (DTCR n )
DMA Timer n Counter Register (DTCN n )
8
32
32
32
Access Reset Value Section/Page
R/W
R/W
R/W
R/W
R/W
R
0x0000
0x00
0x00
0x1111_1111
0x0000_0000
0x0000_0000
24.2.1
DMA Timer Mode Registers (DTMR
n
)
DTMRs, shown in
, program the prescaler and various timer modes.
IPSBAR
Offset:
0x00_0400 (
DTMR0
)
0x00_0440 ( DTMR1 )
0x00_0480 (
DTMR2
)
0x00_04C0 ( DTMR3 )
14 13 15
R
W
Reset 0 0 0
12
PS
11 10 9 8 7
CE
6 5 4 3
Access: User read/write
OM ORRI FRR
0 0 0 0 0 0 0 0
Figure 24-2. DMA Timer Mode Registers (DTMR n )
0 0
2
0
CLK
1
0
0
RST
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
24-3
DMA Timers (DTIM0–DTIM3)
Table 24-2. DTMR n Field Descriptions
Field
15–8
PS
7–6
CE
5
OM
4
ORRI
3
FRR
2–1
CLK
0
RST
Description
Prescaler value. The prescaler is programmed to divide the clock input (internal bus clock/(16 or 1) or clock on
DT n IN) by values from 1 (PS = 0x00) to 256 (PS = 0xFF).
Capture edge.
00 Disable capture event output
01 Capture on rising edge only
10 Capture on falling edge only
11 Capture on any edge
Output mode.
0 Active-low pulse for one internal bus clock cycle (-ns resolution at MHz).
1 Toggle output.
Output reference request, interrupt enable. If ORRI is set when DTER n [REF] = 1, a DMA request or an interrupt occurs, depending on the value of DTXMR n [DMAEN] (DMA request if =1, interrupt if =0).
0 Disable DMA request or interrupt for reference reached (does not affect DMA request or interrupt on capture function).
1 Enable DMA request or interrupt upon reaching the reference value.
Free run/restart
0 Free run. Timer count continues to increment after reaching the reference value.
1 Restart. Timer count is reset immediately after reaching the reference value.
Input clock source for the timer
00 Stop count
01 Internal bus clock divided by 1
10 Internal bus clock divided by 16. Note that this clock source is not synchronized with the timer; thus successive time-outs may vary slightly.
11 DT n IN pin (falling edge)
Reset timer. Performs a software timer reset similar to an external reset, although other register values can still be written while RST = 0. A transition of RST from 1 to 0 resets register values. The timer counter is not clocked unless the timer is enabled.
0 Reset timer (software reset)
1 Enable timer
24.2.2
DMA Timer Extended Mode Registers (DTXMR
n
)
The DTXMR n register programs DMA request and increment modes for the timers.
24-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
DMA Timers (DTIM0–DTIM3)
IPSBAR
Offset:
0x 00_04 02 ( DTXMR0 )
0x
00_044
2
(
DTXMR1
)
0x 00_048 2 ( DTXMR2 )
0x
00_04C
2
(
DTXMR3
)
7 6
0 R
W
R
W
Reset:
DMAEN
DMAEN
0
HALTED
5
0
0
4
0
0
3
0
0
2
0
0
0 0 0 0 0
Figure 24-3. DMA Timer Extended Mode Registers (DTXMR n )
Access: User read/write
1
0
0
0
0
MODE16
MODE16
0
Table 24-3. DTXMR n Field Descriptions
Field Description
7
DMAEN
DMA request. Enables DMA request output on counter reference match or capture edge event.
0 DMA request disabled
1 DMA request enabled
6
HALTED
Controls the counter when the core is stopped. This allows debug mode to be entered without timer interrupts affecting the debug flow.
0 Timer function is not affected by core halt.
1 Timer stops counting while the core is halted.
Note: This bit is only applicable in reference compare mode.
65–1 Reserved, should be cleared.
0
MODE16
Selects the increment mode for the timer. MODE16 = 1 is intended to exercise the upper bits of the 32-bit timer in diagnostic software without requiring the timer to count through its entire dynamic range. When set, the counter’s upper 16 bits mirror its lower 16 bits. All 32 bits of the counter are still compared to the reference value.
0 Increment timer by 1
1 Increment timer by 65,537
24.2.3
DMA Timer Event Registers (DTER
n
)
DTER n , shown in
, reports capture or reference events by setting DTER n [CAP] or
DTER n [REF]. This reporting is done regardless of the corresponding DMA request or interrupt enable values, DTXMR n [DMAEN] and DTMR n [ORRI,CE].
Writing a 1 to either DTER n [REF] or DTER n [CAP] clears it (writing a 0 does not affect bit value); both bits can be cleared at the same time. If configured to generate an interrupt request, the REF and CAP bits should be cleared early in the interrupt service routine so the timer module can negate the interrupt request signal to the interrupt controller. If configured to generate a DMA request, the processing of the DMA data transfer automatically clears both the REF and CAP flags via the internal DMA ACK signal.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
24-5
DMA Timers (DTIM0–DTIM3)
IPSBAR
Offset:
0x00_0403 (DTER0)
0x00_0443 (DTER1)
0x00_0483 (DTER2)
0x00_04C3 (DTER3)
7
0 R
W
Reset: 0
6
0
5
0
4
0
3
0
2
0
0 0 0 0 0
Figure 24-4. DMA Timer Event Registers (DTER n )
Table 24-4. DTER n Field Descriptions
Access: User read/write
1
REF w1c
0
0
CAP w1c
0
Field
7–2
1
REF
Description
Reserved, should be cleared.
Output reference event. The counter value, DTCN n, equals the reference value, DTRR n . Writing a one to REF clears the event condition. Writing a zero has no effect.
REF DTMR n [ORRI] DTXMR n [DMAEN]
1
1
0
1
1
X
0
0
1
1
X
0
1
0
1
No event
No request asserted
No request asserted
Interrupt request asserted
DMA request asserted
0
CAP
Capture event. The counter value has been latched into DTCR n . Writing a one to CAP clears the event condition.
Writing a zero has no effect.
CAP DTMR n [CE]
DTXMR n
[DMAEN]
1
1
1
1
1
1
1
0
1
01
10
10
11
11
XX
00
00
01
1
0
1
0
1
1
0
X
0
No event
Disable capture event output
Disable capture event output
Capture on rising edge & trigger interrupt
Capture on rising edge & trigger DMA
Capture on falling edge & trigger interrupt
Capture on falling edge & trigger DMA
Capture on any edge & trigger interrupt
Capture on any edge & trigger DMA
24-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
DMA Timers (DTIM0–DTIM3)
24.2.4
DMA Timer Reference Registers (DTRR
n
)
Each DTRR n
, shown in Figure 24-5 , contains the reference value compared with the respective
free-running timer counter (DTCN n ) as part of the output-compare function. The reference value is not matched until DTCN n equals DTRR n , and the prescaler indicates that DTCN n should be incremented again. Thus, the reference register is matched after DTRR n +1 time intervals.
IPSBAR
Offset:
0x
00_04
04
(
DTRR0
)
0x 00_044 4 ( DTRR1 )
0x
00_048
4
(
DTRR2
)
0x 00_04C 4 ( DTRR3 )
Access: User read/write
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
R
REF (32-bit reference value)
W
Reset 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
Figure 24-5. DMA Timer Reference Registers (DTRR n )
24.2.5
DMA Timer Capture Registers (DTCR
n
)
Each DTCR n latches the corresponding DTCN n value during a capture operation when an edge occurs on
DT n IN, as programmed in DTMR n . The internal bus clock is assumed to be the clock source. DT n IN cannot simultaneously function as a clocking source and as an input capture pin. Indeterminate operation will result if DT n IN is set as the clock source when the input capture mode is used.
IPSBAR
Offset:
0x 00_04 08 ( DTCR0 )
0x
00_044
8
(
DTCR1
)
0x 00_048 8 ( DTCR2 )
0x
00_04C
8
(
DTCR3
)
Access: User read-only
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
R CAP (32-bit capture counter value)
W
Reset 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Figure 24-6. DMA Timer Capture Registers (DTCR n )
24.2.6
DMA Timer Counters (DTCN
n
)
The current value of the 32-bit DTCNs can be read at anytime without affecting counting. Any write to
DTCN n clears it. The timer counter increments on the clock source rising edge (internal bus clock ÷ 1, internal bus clock ÷ 16, or DT n IN).
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
24-7
DMA Timers (DTIM0–DTIM3)
IPSBAR
Offset:
0x 00_04 0C ( DTCN0 )
0x
00_044
C
(
DTCN1
)
0x 00_048 C ( DTCN2 )
0x
00_04C
C
(
DTCN3
)
Access: User read/write
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
R
W
CNT (32-bit timer counter value count)
Reset 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Figure 24-7. DMA Timer Counters (DTCN n )
24.3
Functional Description
24.3.1
Prescaler
The prescaler clock input is selected from the internal bus clock (f sys
divided by 1 or 16) or from the corresponding timer input, DT n IN. DT n IN is synchronized to the internal bus clock, and the synchronization delay is between two and three internal bus clocks. The corresponding DTMR n [CLK] selects the clock input source. A programmable prescaler divides the clock input by values from 1 to 256.
The prescaler output is an input to the 32-bit counter, DTCN n .
24.3.2
Capture Mode
Each DMA timer has a 32-bit timer capture register (DTCR n ) that latches the counter value when the corresponding input capture edge detector senses a defined DT n IN transition. The capture edge bits
(DTMR n [CE]) select the type of transition that triggers the capture and sets the timer event register capture event bit, DTER n [CAP]. If DTER n [CAP] is set and DTXMR n [DMAEN] is one, a DMA request is asserted. If DTER n [CAP] is set and DTXMR n [DMAEN] is zero, an interrupt is asserted.
24.3.3
Reference Compare
Each DMA timer can be configured to count up to a reference value, at which point DTER n [REF] is set.
If DTMR n [ORRI] is one and DTXMR n [DMAEN] is zero, an interrupt is asserted. If DTMR n [ORRI] is one and DTXMR n [DMAEN] is one, a DMA request is asserted. If the free run/restart bit DTMR n [FRR] is set, a new count starts. If it is clear, the timer keeps running.
24.3.4
Output Mode
When a timer reaches the reference value selected by DTRR, it can send an output signal on DT n OUT.
DT n OUT can be an active-low pulse or a toggle of the current output, as selected by the DTMR n [OM] bit.
24.4
Initialization/Application Information
The general-purpose timer modules typically, but not necessarily, follow this program order:
• The DTMR n and DTXMR n registers are configured for the desired function and behavior.
— Count and compare to a reference value stored in the DTRR n register
24-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
DMA Timers (DTIM0–DTIM3)
— Capture the timer value on an edge detected on DT n IN
— Configure DT n OUT output mode
— Increment counter by 1 or by 65,537 (16-bit mode)
— Enable/disable interrupt or DMA request on counter reference match or capture edge
• The DTMR n [CLK] register is configured to select the clock source to be routed to the prescaler.
— Internal bus clock (can be divided by 1 or 16)
— DT n IN, the maximum value of DT n IN is 1/5 of the internal bus clock, as described in the device’s electrical characteristics
NOTE
DT n IN may not be configured as a clock source when the timer capture mode is selected, or indeterminate operation will result.
• The 8-bit DTMR n [PS] prescaler value is set.
• Using DTMR n [RST], the counter is cleared and started.
• Timer events are either handled with an interrupt service routine, a DMA request, or by a software polling mechanism.
24.4.1
Code Example
The following code provides an example of how to initialize and use DMA Timer0 for counting time-out periods.
DTMR0 EQU IPSBARx+0x400 ;Timer0 mode register
DTMR1 EQU IPSBARx+0x440 ;Timer1 mode register
DTRR0 EQU IPSBARx+0x404 ;Timer0 reference register
DTRR1 EQU IPSBARx+0x444 ;Timer1 reference register
DTCR0 EQU IPSBARx+0x408 ;Timer0 capture register
DTCR1 EQU IPSBARx+0x448 ;Timer1 capture register
DTCN0 EQU IPSBARx+0x40C ;Timer0 counter register
DTCN1 EQU IPSBARx+0x44C ;Timer1 counter register
DTER0 EQU IPSBARx+0x403 ;Timer0 event register
DTER1 EQU IPSBARx+0x443 ;Timer1 event register
* TMR0 is defined as: *
*[PS] = 0xFF, divide clock by 256
*[CE] = 00
*[OM] = 0 disable capture event output output=active-low pulse
*[ORRI] = 0,
*[FRR] = 1,
*[CLK] = 10,
*[RST] = 0, disable ref. match output restart mode enabled internal bus clock/16 timer0 disabled move.w #0xFF0C,D0 move.w D0,TMR0 move.l #0x0000,D0;writing to the timer counter with any move.l DO,TCN0 ;value resets it to zero move.l #0xAFAF,DO ;set the timer0 reference to be move.l #D0,TRR0 ;defined as 0xAFAF
The simple example below uses Timer0 to count time-out loops. A time-out occurs when the reference value, 0xAFAF, is reached.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
24-9
DMA Timers (DTIM0–DTIM3) timer0_ex clr.l DO clr.l D1 clr.l D2 move.l #0x0000,D0 move.l D0,TCN0 move.b #0x03,D0 move.b D0,TER0 move.w TMR0,D0 bset #0,D0 move.w D0,TMR0
T0_LOOP move.b TER0,D1 btst #1,D1 beq T0_LOOP addi.l #1,D2 cmp.l #5,D2 beq T0_FINISH move.b #0x02,D0 move.b D0,TER0 jmp T0_LOOP
T0_FINISH
HALT
;reset the counter to 0x0000
;writing ones to TER0[REF,CAP]
;clears the event flags
;save the contents of TMR0 while setting
;the 0 bit. This enables timer 0 and starts counting
;load the value back into the register, setting TMR0[RST]
;load TER0 and see if
;TER0[REF] has been set
;Increment D2
;Did D2 reach 5? (i.e. timer ref has timed)
;If so, end timer0 example. Otherwise jump back.
;writing one to TER0[REF] clears the event flag
;End processing. Example is finished
24.4.2
Calculating Time-Out Values
The formula below determines time-out periods for various reference values:
Timeout period = ( ) × ( 1 or 16 ) × ( DTMR n [PS] 1 ) × ( DTRR n [REF] 1 )
When calculating time-out periods, add 1 to the prescaler to simplify calculating, because
DTMR n [PS] = 0x00 yields a prescaler of 1, and DTMR n [PS] = 0xFF yields a prescaler of 256.
Eqn. 24-1
For example, if a 66-MHz timer clock is divided by 16, DTMR n [PS] = 0x7F, and the timer is referenced at 0xFBC5 (64453 decimal), the time-out period is as follows:
Timeout period =
1
6
× 16 × ( + ) × ( 64453 1 ) = 2.00s
Eqn. 24-2
24-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 25
Queued Serial Peripheral Interface (QSPI)
25.1
Introduction
This chapter describes the queued serial peripheral interface (QSPI) module. Following a feature set overview is a description of operation including details of the QSPI’s internal RAM organization. The chapter concludes with the programming model and a timing diagram.
25.1.1
Block Diagram
illustrates the QSPI module.
Queue Control
Block
Queue
Pointer
4
Comparator
80-byte
QSPI
RAM
End Queue
Pointer
Done
QSPI
Address
Register
4
QSPI
Data
Register
Control Logic
Status
Regs
Control
Regs
Chip
Selects
Logic
Array 4
Command msb lsb
8/16 Bit Shift Reg.
Rx/Tx Data Reg.
QSPI_DIN
QSPI_DOUT
QSPI_CS[:0]
Delay
Counter
Internal Bus
Internal Bus
Clock (f sys
)
Divide by 2
Baud Rate
Generator
Figure 25-1. QSPI Block Diagram
QSPI_CLK
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
25-1
Queued Serial Peripheral Interface (QSPI)
25.1.2
Overview
The queued serial peripheral interface module provides a serial peripheral interface with queued transfer capability. It allows users to queue up to 16 transfers at once, eliminating CPU intervention between transfers. Transfer RAM in the QSPI is indirectly accessible using address and data registers.
NOTE
The GPIO module must be configured to enable the peripheral function of the appropriate pins (refer to Chapter 13, “General Purpose I/O Module” ) prior to configuring the QSPI Module.
25.1.3
Features
Features include the following:
• Programmable queue to support up to 16 transfers without user intervention
• Supports transfer sizes of 8 to 16 bits in 1-bit increments
• Four peripheral chip-select lines for control of up to 15 devices (All four chip selects may not be available on all devices. See Chapter 2, “Signal Descriptions,” for details on which chip-selects are pinned-out.)
• Baud rates from 129.4 Kbps to 16.6 Mbps at 66 MHz internal bus frequency
• Programmable delays before and after transfers
• Programmable QSPI clock phase and polarity
• Supports wraparound mode for continuous transfers
25.1.4
Modes of Operation
Because the QSPI module only operates in master mode, the master bit in the QSPI mode register
(QMR[MSTR]) must be set for the QSPI to function properly. If the master bit is not set, QSPI activity will be indeterminate. The QSPI can initiate serial transfers but cannot respond to transfers initiated by other QSPI masters.
25.2
External Signal Description
The module provides access to as many as 15 devices with a total of seven signals: QSPI_DOUT,
QSPI_DIN, QSPI_CLK, QSPI_CS0, QSPI_CS1, and QSPI_CS2, and QSPI_CS3.
Peripheral chip-select signals, QSPI_CS n , are used to select an external device as the source or destination for serial data transfer. Signals are asserted whenever a command in the queue is executed. More than one chip-select signal can be asserted simultaneously.
Although QSPI_CS n will function as simple chip selects in most applications, up to 15 devices can be selected by decoding them with an external 4-to-16 decoder.
25-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Queued Serial Peripheral Interface (QSPI)
Table 25-1. QSPI Input and Output Signals and Functions
Signal Name
QSPI Data Output (QSPI_DOUT)
QSPI Data Input (QSPI_DIN)
Serial Clock (QSPI_CLK)
Peripheral Chip Selects (QSPI_CS n )
Hi-Z or Actively Driven
Configurable
N/A
Actively driven
Actively driven
Function
Serial data output from QSPI
Serial data input to QSPI
Clock output from QSPI
Peripheral selects
25.3
Memory Map/Register Definition
is the QSPI register memory map. Reading reserved locations returns zeros.
Table 25-2. QSPI Memory Map
IPSBAR
Offset
1
0x00_0340
0x00_0344
0x00_0348
0x00_034C
Register
QSPI Mode Register (QMR)
QSPI Delay Register (QDLYR)
QSPI Wrap Register (QWR)
QSPI Interrupt Register (QIR)
Width
(bits)
Access Reset Value Section/Page
16
16
16
16
R/W
R/W
R/W
2
R/W
2
R/W
2
0x00_0350 QSPI Address Register (QAR) 16
0x00_0354 QSPI Data Register (QDR) 16 R/W
1
2
Addresses not assigned to a register and undefined register bits are reserved for expansion.
See the register description for special cases. Some bits may be read- or write-only.
0x0104
0x0404
0x0000
0x0000
0x0000
0x0000
25.3.1
QSPI Mode Register (QMR)
The QMR, shown in
, determines the basic operating modes of the QSPI module. Parameters such as QSPI_CLK polarity and phase, baud rate, master mode operation, and transfer size are determined by this register. The data output high impedance enable, DOHIE, controls the operation of QSPI_DOUT between data transfers. When DOHIE is cleared, QSPI_DOUT is actively driven between transfers. When
DOHIE is set, QSPI_DOUT assumes a high impedance state.
NOTE
Because the QSPI does not operate in slave mode, the master mode enable bit (QMR[MSTR]) must be set for the QSPI module to operate correctly.
Access: User read/write IPSBAR
Offset:
0x00_0340
15 14
R
MSTR DOHIE
W
Reset 0 0
13
0
12
BITS
11 10 9 8
CPOL CPHA
7 6 5
0 0 0 0 1 0 0
Figure 25-2. QSPI Mode Register (QMR)
0
4
BAUD
3
0 0
2
1
1
0
0
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
25-3
Queued Serial Peripheral Interface (QSPI)
Table 25-3. QMR Field Descriptions
Field Description
15
MSTR
14
DOHIE
13–10
BITS
Master mode enable.
0 Reserved, do not use.
1 The QSPI is in master mode. Must be set for the QSPI module to operate correctly.
Data output high impedance enable. Selects QSPI_DOUT mode of operation.
0 Default value after reset. QSPI_DOUT is actively driven between transfers.
1 QSPI_DOUT is high impedance between transfers.
Transfer size. Determines the number of bits to be transferred for each entry in the queue.
BITS Bits per Transfer
0000
0001–0111
1000
1001
1010
1011
1100
1101
1110
1111
16
Reserved
8
9
10
11
12
13
14
15
9
CPOL
8
CPHA
7–0
BAUD
Clock polarity. Defines the clock polarity of QSPI_CLK.
0 The inactive state value of QSPI_CLK is logic level 0.
1 The inactive state value of QSPI_CLK is logic level 1.
Clock phase. Defines the QSPI_CLK clock-phase.
0 Data captured on the leading edge of QSPI_CLK and changed on the following edge of QSPI_CLK.
1 Data changed on the leading edge of QSPI_CLK and captured on the following edge of QSPI_CLK.
Baud rate divider. The baud rate is selected by writing a value in the range 2–255. A value of zero disables the QSPI.
A value of 1 is an invalid setting. The desired QSPI_CLK baud rate is related to the internal bus clock and
QMR[BAUD] by the following expression:
QMR[BAUD] = f sys/
/ (2 ×
[desired QSPI_CLK baud rate])
25-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
shows an example of a QSPI clocking and data transfer.
QSPI_CLK
Queued Serial Peripheral Interface (QSPI)
QSPI_DOUT 15 14 13 12 11 10 msb
9 8 7 6 5 4 3 2 1 0
QSPI_DIN 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
A
B
QSPI_CS
QMR[CPOL] = 0
QMR[CPHA] = 1
QCR[CONT] = 0
Chip selects are active low
A = QDLYR[QCD]
B = QDLYR[DTL]
Figure 25-3. QSPI Clocking and Data Transfer Example
25.3.2
QSPI Delay Register (QDLYR)
IPSBAR
Offset:
0x00_0344
14 15
R
SPE
W
Reset 0 0
13
0
12 11
QCD
10 9 8 7 6 5
0 0 1 0 0 0 0 0
Figure 25-4. QSPI Delay Register (QDLYR)
Table 25-4. QDLYR Field Descriptions
4
0
DTL
3
0
Access: User read/write
2
1
1
0
0
0
Field
15
SPE
14–8
QCD
7–0
DTL
Description
QSPI enable. When set, the QSPI initiates transfers in master mode by executing commands in the command RAM.
Automatically cleared by the QSPI when a transfer completes. The user can also clear this bit to abort transfer unless
QIR[ABRTL] is set. The recommended method for aborting transfers is to set QWR[HALT].
QSPI_CLK delay. When the DSCK bit in the command RAM is set this field determines the length of the delay from assertion of the chip selects to valid QSPI_CLK transition. See
Section 25.4.3, “Transfer Delays”
for information on setting this bit field.
Delay after transfer. When the DT bit in the command RAM is set this field determines the length of delay after the serial transfer.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
25-5
Queued Serial Peripheral Interface (QSPI)
25.3.3
QSPI Wrap Register (QWR)
IPSBAR
Offset:
0x00_0348
15 14 13 12
R
HALT WREN WRTO CSIV
W
Reset 0 0 0 0
11 10 9
ENDQP
8 7 6 5
CPTQP
0 0 0 0 0 0
Figure 25-5. QSPI Wrap Register (QWR)
0
4
0
3
0
Access: User read/write
2 1
NEWQP
0 0
0
0
Table 25-5. QWR Field Descriptions
Field Description
15
HALT
14
WREN
13
WRTO
12
CSIV
Halt transfers. Assertion of this bit causes the QSPI to stop execution of commands once it has completed execution of the current command.
Wraparound enable. Enables wraparound mode.
0 Execution stops after executing the command pointed to by QWR[ENDQP].
1 After executing command pointed to by QWR[ENDQP], wrap back to entry zero, or the entry pointed to by
QWR[NEWQP] and continue execution.
Wraparound location. Determines where the QSPI wraps to in wraparound mode.
0 Wrap to RAM entry zero.
1 Wrap to RAM entry pointed to by QWR[NEWQP].
QSPI_CS inactive level.
0 QSPI chip select outputs return to zero when not driven from the value in the current command RAM entry during a transfer (that is, inactive state is 0, chip selects are active high).
1 QSPI chip select outputs return to one when not driven from the value in the current command RAM entry during a transfer (that is, inactive state is 1, chip selects are active low).
11–8
ENDQP
End of queue pointer. Points to the RAM entry that contains the last transfer description in the queue.
7–4
CPTQP
Completed queue entry pointer. Points to the RAM entry that contains the last command to have been completed.
This field is read only.
3–0
NEWQP
Start of queue pointer. This 4-bit field points to the first entry in the RAM to be executed on initiating a transfer.
25.3.4
QSPI Interrupt Register (QIR)
IPSBAR
Offset:
0x00_034C
15 14
R
WCEFB ABRTB
W
Reset 0 0
13
0
0
12
0
11
0
10
ABRTL WCEFE ABRTE
0
9
0
0
8
SPIFE
0
7
0
0
6
0
0
Figure 25-6. QSPI Interrupt Register (QIR)
0
5
0
Access: User read/write
4 3 2 1 0
0 WCEF ABRT 0 SPIF
0 w1c w1c
0 0 0 w1c
0
25-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Queued Serial Peripheral Interface (QSPI)
Table 25-6. QIR Field Descriptions
Field Description
15
WCEFB
Write collision access error enable. A write collision occurs during a data transfer when the RAM entry containing the current command is written to by the CPU with the QDR. When this bit is asserted, the write access to QDR results in an access error.
14
ABRTB
Abort access error enable. An abort occurs when QDLYR[SPE] is cleared during a transfer. When set, an attempt to clear QDLYR[SPE] during a transfer results in an access error.
13
12
ABRTL
Reserved, should be cleared.
Abort lock-out. When set, QDLYR[SPE] cannot be cleared by writing to the QDLYR. QDLYR[SPE] is only cleared by the QSPI when a transfer completes.
7–4
3
WCEF
2
ABRT
1
0
SPIF
11
WCEFE
Write collision (WCEF) interrupt enable.
0 Write collision interrupt disabled
1 Write collision interrupt enabled
10
ABRTE
Abort (ABRT) interrupt enable.
0 Abort interrupt disabled
1 Abort interrupt enabled
9
8
SPIFE
Reserved, should be cleared.
QSPI finished (SPIF) interrupt enable.
0 SPIF interrupt disabled
1 SPIF interrupt enabled
Reserved, should be cleared.
Write collision error flag. Indicates that an attempt has been made to write to the RAM entry that is currently being executed. Writing a 1 to this bit (w1c) clears it and writing 0 has no effect.
Abort flag. Indicates that QDLYR[SPE] has been cleared by the user writing to the QDLYR rather than by completion of the command queue by the QSPI. Writing a 1 to this bit (w1c) clears it and writing 0 has no effect.
Reserved, should be cleared.
QSPI finished flag. Asserted when the QSPI has completed all the commands in the queue. Set on completion of the command pointed to by QWR[ENDQP], and on completion of the current command after assertion of
QWR[HALT]. In wraparound mode, this bit is set every time the command pointed to by QWR[ENDQP] is completed.
Writing a 1 to this bit (w1c) clears it and writing 0 has no effect.
The command and data RAM in the QSPI are indirectly accessible with QDR and QAR as 48 separate locations that comprise 16 words of transmit data, 16 words of receive data, and 16 bytes of commands.
A write to QDR causes data to be written to the RAM entry specified by QAR[ADDR]. This also causes the value in QAR to increment.
Correspondingly, a read at QDR returns the data in the RAM at the address specified by QAR[ADDR].
This also causes QAR to increment. A read access requires a single wait state.
NOTE
The QAR does not wrap after the last queue entry within each section of the
RAM. The application software must handle address range errors.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
25-7
Queued Serial Peripheral Interface (QSPI)
25.3.5
QSPI Address Register (QAR)
The QAR is used to specify the location in the QSPI RAM that read and write operations affect. As shown
, the transmit RAM is located at addresses 0x0 to 0xF, the receive RAM is located at 0x10 to 0x1F, and the command RAM is located at 0x20 to 0x2F. These addresses refer to the
QSPI RAM space, not the device memory map.
Access: User read/write IPSBAR
Offset:
0x0x00_0350
15
R 0
W
Reset 0
14
0
0
13
0
0
12
0
11
0
10
0
9
0
8
0
7
0
6
0
5
0
4
0 0 0
1
0
0
0 0 0 0 0 0 0 0
Figure 25-7. QSPI Address Register (QAR)
3
ADDR
2
25.3.6
QSPI Data Register (QDR)
The QDR, shown in Figure 25-8 , is used to access QSPI RAM indirectly. The CPU reads and writes all
data from and to the QSPI RAM through this register.
Access: User read/write IPSBAR
Offset:
0x0x00_0354
14 15
R
W
Reset 0 0
13
0
12
0
11 10 9 8
DATA
7 6 5
0 0 0 0 0 0
Figure 25-8. QSPI Data Register (QDR)
0
4
0
3
0
2
0
1
0
0
0
25.3.7
Command RAM Registers (QCR0–QCR15)
The command RAM is accessed using the upper byte of the QDR; the QSPI cannot modify information in command RAM. There are 16 bytes in the command RAM. Each byte is divided into two fields. The chip select field enables external peripherals for transfer. The command field provides transfer operations.
NOTE
The command RAM is accessed only using the most significant byte of
QDR and indirect addressing based on QAR[ADDR].
Address: QAR[ADDR] Access: CPU write-only
15 14 13 12
R
W CONT BITSE DT DSCK
Reset — — — —
11
—
10 9
QSPI_CS
— —
8
—
7 6 5 4 3 2 1 0
0 0 0 0 0 0 0 0
— — — — — — — —
Figure 25-9. Command RAM Registers (QCR0–QCR15)
25-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Queued Serial Peripheral Interface (QSPI)
Table 25-7. QCR0–QCR15 Field Descriptions
Field Description
15
CONT
14
BITSE
13
DT
Continuous.
0 Chip selects return to inactive level defined by QWR[CSIV] when a single word transfer is complete.
1 Chip selects return to inactive level defined by QWR[CSIV] only after the transfer of the queue entries (max of 16 words).
Note: In order to keep the chip selects asserted for transfers beyond 16 words, the QWR[CSIV] bit must be set to control the level that the chip selects return to after the first transfer.
Bits per transfer enable.
0 Eight bits
1 Number of bits set in QMR[BITS]
Delay after transfer enable.
0 Default reset value.
1 The QSPI provides a variable delay at the end of serial transfer to facilitate interfacing with peripherals that have a latency requirement. The delay between transfers is determined by QDLYR[DTL].
12
DSCK
Chip select to QSPI_CLK delay enable.
0 Chip select valid to QSPI_CLK transition is one-half QSPI_CLK period.
1 QDLYR[QCD] specifies the delay from QSPI_CS valid to QSPI_CLK.
11–8
QSPI_CS
Peripheral chip selects. Used to select an external device for serial data transfer. More than one chip select may be active at once, and more than one device can be connected to each chip select. Bits 11-8 map directly to the corresponding QSPI_CS n pins. If more than four chip selects are needed, then an external demultiplexor can be used with the QSPI_CS n pins.
Note: Not all chip selects may be available on all device packages. See Chapter 2, “Signal Descriptions,” for details on which chip selects are pinned-out.
7–0 Reserved, should be cleared.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
25-9
Queued Serial Peripheral Interface (QSPI)
QSPI_CS n
(active-low)
QSPI_CLK
QS1
QS2
QSPI_DOUT
QS3 QS4 QS5
QSPI_DIN
QS1: QSPI_CS to QSPI_CLK
QS2: QSPI_CLK to QSPI_DOUT VALID
QS3: QSPI_CLK to QSPI_DOUT HOLD
QS4: QSPI_DIN to QSPI_CLK SETUP
QS5: QSPI_DIN to QSPI_CLK HOLD
1T1 is defined as the clock period in ns.
Min
1T1
0 ns
10 ns
10 ns
Max
20 ns
Figure 25-10. QSPI Timing
25.4
Functional Description
The QSPI uses a dedicated 80-byte block of static RAM accessible to both the module and CPU to perform queued operations. The RAM is divided into three segments:
• 16 command control bytes (command RAM)
• 32 transmit data bytes (transfer RAM)
• 32 receive data bytes (transfer RAM)
The RAM is organized so that 1 byte of command control data, 1 word of transmit data, and 1 word of receive data comprise 1 of the 16 queue entries (0x0–0xF).
NOTE
Throughout ColdFire documentation, the term ‘word’ is used to designate a
16-bit data unit. The only exceptions to this appear in discussions of serial communication modules such as QSPI that support variable-length data units. To simplify these discussions, the functional unit is referred to as a
‘word’ regardless of length.
25-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Queued Serial Peripheral Interface (QSPI)
The user initiates QSPI operation by loading a queue of commands in command RAM, writing transmit data into transmit RAM, and then enabling the QSPI data transfer. The QSPI executes the queued commands and sets the completion flag in the QSPI interrupt register (QIR[SPIF]) to signal their completion. As another option, QIR[SPIFE] can be enabled to generate an interrupt.
The QSPI uses four queue pointers. The user can access three of them through fields in QSPI wrap register
(QWR):
• New queue pointer (QWR[NEWQP])—points to the first command in the queue
• Internal queue pointer—points to the command currently being executed
• Completed queue pointer (QWR[CPTQP])—points to the last command executed
• End queue pointer (QWR[ENDQP]) —points to the final command in the queue
The internal pointer is initialized to the same value as QWR[NEWQP]. During normal operation, the following sequence repeats:
1. The command pointed to by the internal pointer is executed.
2. The value in the internal pointer is copied into QWR[CPTQP].
3. The internal pointer is incremented.
Execution continues at the internal pointer address unless the QWR[NEWQP] value is changed. After each command is executed, QWR[ENDQP] and QWR[CPTQP] are compared. When a match occurs,
QIR[SPIF] is set and the QSPI stops unless wraparound mode is enabled. Setting QWR[WREN] enables wraparound mode.
QWR[NEWQP] is cleared at reset. When the QSPI is enabled, execution begins at address 0x0 unless another value has been written into QWR[NEWQP]. QWR[ENDQP] is cleared at reset but is changed to show the last queue entry before the QSPI is enabled. QWR[NEWQP] and QWR[ENDQP] can be written at any time. When the QWR[NEWQP] value changes, the internal pointer value also changes unless a transfer is in progress, in which case the transfer completes normally. Leaving QWR[NEWQP] and
QWR[ENDQP] set to 0x0 causes a single transfer to occur when the QSPI is enabled.
Data is transferred relative to QSPI_CLK, which can be generated in any one of four combinations of phase and polarity using QMR[CPHA,CPOL]. Data is transferred with the most significant bit (msb) first.
The number of bits transferred defaults to 8, but can be set to any value between 8 and 16 by writing a value into the BITSE field of the command RAM (QCR[BITSE]).
25.4.1
QSPI RAM
The QSPI contains an 80-byte block of static RAM that can be accessed by both the user and the QSPI.
This RAM does not appear in the device memory map, because it can only be accessed by the user indirectly through the QSPI address register (QAR) and the QSPI data register (QDR). The RAM is divided into three segments with 16 addresses each:
• Receive data RAM—the initial destination for all incoming data
• Transmit data RAM—a buffer for all out-bound data
• Command RAM—where commands are loaded
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
25-11
Queued Serial Peripheral Interface (QSPI)
The transmit and command RAM are user write-only. The receive RAM is user read-only. Figure 25-11
shows the RAM configuration. The RAM contents are undefined immediately after a reset.
The command and data RAM in the QSPI is indirectly accessible with QDR and QAR as 48 separate locations that comprise 16 words of transmit data, 16 words of receive data, and 16 bytes of commands.
A write to QDR causes data to be written to the RAM entry specified by QAR[ADDR] and causes the value in QAR to increment. Correspondingly, a read at QDR returns the data in the RAM at the address specified by QAR[ADDR]. This also causes QAR to increment. A read access requires a single wait state.
Relative
Address
0x00
0x01
...
0x0F
Register Function
QTR0 Transmit RAM
QTR1
...
QTR15
16 bits wide
0x10
0x11
...
0x1F
QRR0 Receive RAM
QRR1
...
QRR15
16 bits wide
0x20
0x21
...
0x2F
QCR0 Command RAM
QCR1
...
QCR15
8 bits wide
Figure 25-11. QSPI RAM Model
25.4.1.1
Receive RAM
Data received by the QSPI is stored in the receive RAM segment located at 0x10 to 0x1F in the QSPI RAM space. The user reads this segment to retrieve data from the QSPI. Data words with less than 16 bits are stored in the least significant bits of the RAM. Unused bits in a receive queue entry are set to zero upon completion of the individual queue entry.
QWR[CPTQP] shows which queue entries have been executed. The user can query this field to determine which locations in receive RAM contain valid data.
25.4.1.2
Transmit RAM
Data to be transmitted by the QSPI is stored in the transmit RAM segment located at addresses 0x0 to 0xF.
The user normally writes 1 word into this segment for each queue command to be executed. The user cannot read data in the transmit RAM.
25-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Queued Serial Peripheral Interface (QSPI)
Outbound data must be written to transmit RAM in a right-justified format. The unused bits are ignored.
The QSPI copies the data to its data serializer (shift register) for transmission. The data is transmitted most significant bit first and remains in transmit RAM until overwritten by the user.
25.4.1.3
Command RAM
The CPU writes one byte of control information to this segment for each QSPI command to be executed.
Command RAM, referred to as QCR0–15, is write-only memory from a user’s perspective.
Command RAM consists of 16 bytes, each divided into two fields. The peripheral chip select field controls the QSPI_CS signal levels for the transfer. The command control field provides transfer options.
A maximum of 16 commands can be in the queue. Queue execution proceeds from the address in
QWR[NEWQP] through the address in QWR[ENDQP].
The QSPI executes a queue of commands defined by the control bits in each command RAM entry that sequence the following actions:
• Chip-select pins are activated.
• Data is transmitted from transmit RAM and received into the receive RAM.
• The synchronous transfer clock QSPI_CLK is generated.
Before any data transfers begin, control data must be written to the command RAM, and any out-bound data must be written to transmit RAM. Also, the queue pointers must be initialized to the first and last entries in the command queue.
Data transfer is synchronized with the internally generated QSPI_CLK, whose phase and polarity are controlled by QMR[CPHA] and QMR[CPOL]. These control bits determine which QSPI_CLK edge is used to drive outgoing data and to latch incoming data.
25.4.2
Baud Rate Selection
The maximum QSPI clock frequency is one-fourth the clock frequency of the internal bus clock (f sys
).
Baud rate is selected by writing a value from 2–255 into QMR[BAUD]. The QSPI uses a prescaler to derive the QSPI_CLK rate from the internal bus clock divided by two.
A baud rate value of zero turns off the QSPI_CLK.
The desired QSPI_CLK baud rate is related to the internal bus clock and QMR[BAUD] by the following expression:
QMR[BAUD] = f
----------------------------------------------------------------------------------Eqn. 25-1
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
25-13
Queued Serial Peripheral Interface (QSPI)
Table 25-8. QSPI_CLK Frequency as Function of Internal Bus Clock and Baud Rate
Internal Bus Clock = 66 MHz
QMR [BAUD]
8
16
2
4
32
255
QSPI_CLK
16.5 MHz
8.25 MHz
4.1 MHz
2.06 MHz
1.0 MHz
12.9 kHz
25.4.3
Transfer Delays
The QSPI supports programmable delays for the QSPI_CS signals before and after a transfer. The time between QSPI_CS assertion and the leading QSPI_CLK edge, and the time between the end of one transfer and the beginning of the next, are both independently programmable.
The chip select to clock delay enable bit in command RAM, QCR[DSCK], enables the programmable delay period from QSPI_CS assertion until the leading edge of QSPI_CLK. QDLYR[QCD] determines the period of delay before the leading edge of QSPI_CLK. The following expression determines the actual delay before the QSPI_CLK leading edge:
QSPI_CS-to-QSPI_CLK delay =
QDLYR[QCD] has a range of 1–127.
f sys
Eqn. 25-2
When QDLYR[QCD] or QCR[DSCK] equals zero, the standard delay of one-half the QSPI_CLK period is used.
The command RAM delay after transmit enable bit, QCR[DT], enables the programmable delay period from the negation of the QSPI_CS signals until the start of the next transfer. The delay after transfer can be used to provide a peripheral deselect interval. A delay can also be inserted between consecutive transfers to allow serial A/D converters to complete conversion. There are two transfer delay options: the user can choose to delay a standard period after serial transfer is complete or can specify a delay period.
Writing a value to QDLYR[DTL] specifies a delay period. QCR[DT] determines whether the standard delay period (DT = 0) or the specified delay period (DT = 1) is used. The following expression is used to calculate the delay when DT = 1:
Delay after transfer =
32 QDLYR[DTL] f sys
(DT = 1) Eqn. 25-3 where QDLYR[DTL] has a range of 1–255. A zero value for DTL causes a delay-after-transfer value of
8192/f sys/
. Standard delay period (DT = 0) is calculated by the following:
Standard delay after transfer = f
17
(DT = 0) sys
Eqn. 25-4
Adequate delay between transfers must be specified for long data streams because the QSPI module requires time to load a transmit RAM entry for transfer. Receiving devices need at least the standard delay
25-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Queued Serial Peripheral Interface (QSPI) between successive transfers. If the internal bus clock is operating at a slower rate, the delay between transfers must be increased proportionately.
25.4.4
Transfer Length
There are two transfer length options. The user can choose a default value of 8 bits or a programmed value of 8 to 16 bits. The programmed value must be written into QMR[BITS]. The command RAM bits per transfer enable field, QCR[BITSE], determines whether the default value (BITSE = 0) or the BITS[3–0] value (BITSE = 1) is used. QMR[BITS] indicates the required number of bits to be transferred, with the default value of 16 bits.
25.4.5
Data Transfer
The transfer operation is initiated by setting QDLYR[SPE]. Shortly after QDLYR[SPE] is set, the QSPI executes the command at the command RAM address pointed to by QWR[NEWQP]. Data at the pointer address in transmit RAM is loaded into the data serializer and transmitted. Data that is simultaneously received is stored at the pointer address in receive RAM.
When the proper number of bits has been transferred, the QSPI stores the working queue pointer value in
QWR[CPTQP], increments the working queue pointer, and loads the next data for transfer from the transmit RAM. The command pointed to by the incremented working queue pointer is executed next unless a new value has been written to QWR[NEWQP]. If a new queue pointer value is written while a transfer is in progress, the current transfer is completed normally.
When the CONT bit in the command RAM is set, the QSPI_CS signals are asserted between transfers.
When CONT is cleared, QSPI_CS n are negated between transfers. Note, the QSPI_CS signals are not high impedance.
When the QSPI reaches the end of the queue, it asserts the SPIF flag, QIR[SPIF]. If QIR[SPIFE] is set, an interrupt request is generated when QIR[SPIF] is asserted. Then the QSPI clears QDLYR[SPE] and stops, unless wraparound mode is enabled.
Wraparound mode is enabled by setting QWR[WREN]. The queue can wrap to pointer address 0x0, or to the address specified by QWR[NEWQP], depending on the state of QWR[WRTO].
In wraparound mode, the QSPI cycles through the queue continuously, even while requesting interrupt service. QDLYR[SPE] is not cleared when the last command in the queue is executed. New receive data overwrites previously received data in the receive RAM. Each time the end of the queue is reached,
QIR[SPIFE] is set. QIR[SPIF] is not automatically reset. If interrupt driven QSPI service is used, the service routine must clear QIR[SPIF] to abort the current request. Additional interrupt requests during servicing can be prevented by clearing QIR[SPIFE].
There are two recommended methods of exiting wraparound mode: clearing QWR[WREN] or setting
QWR[HALT]. Exiting wraparound mode by clearing QDLYR[SPE] is not recommended because this may abort a serial transfer in progress. The QSPI sets SPIF, clears QDLYR[SPE], and stops the first time it reaches the end of the queue after QWR[WREN] is cleared. After QWR[HALT] is set, the QSPI finishes the current transfer, then stops executing commands. After the QSPI stops, QDLYR[SPE] can be cleared.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
25-15
Queued Serial Peripheral Interface (QSPI)
25.5
Initialization/Application Information
The following steps are necessary to set up the QSPI 12-bit data transfers and a QSPI_CLK of 4.125 MHz.
The QSPI RAM is set up for a queue of 16 transfers. All four QSPI_CS signals are used in this example.
1. Write the QMR with 0xB308 to set up 12-bit data words with the data shifted on the falling clock edge, and a QSPI_CLK frequency of 4.125 MHz (assuming a 66-MHz internal bus clock).
2. Write QDLYR with the desired delays.
3. Write QIR with 0xD00F to enable write collision, abort bus errors, and clear any interrupts.
4. Write QAR with 0x0020 to select the first command RAM entry.
5. Write QDR with 0x7E00, 0x7E00, 0x7E00, 0x7E00, 0x7D00, 0x7D00, 0x7D00, 0x7D00, 0x7B00,
0x7B00, 0x7B00, 0x7B00, 0x7700, 0x7700, 0x7700, and 0x7700 to set up four transfers for each chip select. The chip selects are active low in this example.
6. Write QAR with 0x0000 to select the first transmit RAM entry.
7. Write QDR with sixteen 12-bit words of data.
8. Write QWR with 0x0F00 to set up a queue beginning at entry 0 and ending at entry 15.
9. Set QDLYR[SPE] to enable the transfers.
10. Wait until the transfers are complete. QIR[SPIF] is set when the transfers are complete.
11. Write QAR with 0x0010 to select the first receive RAM entry.
12. Read QDR to get the received data for each transfer.
13. Repeat steps 5 through 13 to do another transfer.
25-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 26
UART Modules
26.1
Introduction
This chapter describes the use of the three universal asynchronous receiver/transmitters (UARTs) and includes programming examples.
NOTE
The designation ‘ n ’ is used throughout this section to refer to registers or signals associated with one of the three identical UART modules: UART0,
UART1, or UART2.
26.1.1
Overview
Each of the three independent UARTs can be clocked by the internal bus clock, eliminating the need for an external UART clock. As
shows, each UART module interfaces directly to the CPU and consists of the following:
• Serial communication channel
• Programmable clock generation
• Interrupt control logic and DMA request logic
• Internal channel control logic
UART
Internal Channel
Control Logic
Serial
Communications
Channel
U n CTS
U n RTS
U n RXD
U n TXD
Interrupt Request
(to Interrupt Controller)
Interrupt Control
Logic
Transmit DMA Request
Receive DMA Request
(To DMA Controller)
DMA Request
Logic
Programmable
Clock
Generation
Internal Bus Clock or External clock (DT n IN)
Figure 26-1. UART Block Diagram
NOTE
UART n can be clocked by the DT n IN pin. However, if the timers are used, then input capture mode is not available for that timer.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
26-1
UART Modules
The serial communication channel provides a full-duplex asynchronous/synchronous receiver and transmitter deriving an operating frequency from the internal bus clock or an external clock using the timer pin. The transmitter converts parallel data from the CPU to a serial bit stream, inserting appropriate start, stop, and parity bits. It outputs the resulting stream on the transmitter serial data output (U n TXD). See
Section 26.4.2.1, “Transmitter
.”
The receiver converts serial data from the receiver serial data input (U n RXD) to parallel format, checks for a start, stop, and parity bits, or break conditions, and transfers the assembled character onto the bus during read operations. The receiver may be polled, interrupt driven, or use DMA requests for servicing.
See Section 26.4.2.2, “Receiver
.”
NOTE
The GPIO module must be configured to enable the peripheral function of the appropriate pins (refer to Chapter 13, “General Purpose I/O Module” ) prior to configuring the UART module.
26.1.2
Features
The device contains three independent UART modules with the following features:
• Each clocked by an external clock or by the internal bus clock (eliminating a need for an external
UART clock)
• Full-duplex asynchronous/synchronous receiver/transmitter
• Quadruple-buffered receiver
• Double-buffered transmitter
• Independently programmable receiver and transmitter clock sources
• Programmable data format:
— 5–8 data bits plus parity
— Odd, even, no parity, or force parity
— One, one-and-a-half, or two stop bits
• Each serial channel programmable to normal (full-duplex), automatic echo, local loop-back, or remote loop-back mode
• Automatic wake-up mode for multidrop applications
• Four maskable interrupt conditions
• All three UARTs have DMA request capability
• Parity, framing, and overrun error detection
• False-start bit detection
• Line-break detection and generation
• Detection of breaks originating in the middle of a character
• Start/end break interrupt/status
26-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
UART Modules
26.2
External Signal Description
shows both the external and internal signal groups.
An internal interrupt request signal is provided to notify the interrupt controller of an interrupt condition.
The output is the logical NOR of unmasked UISR n bits. The interrupt level and priority are programmed in the interrupt controller. See Chapter 14, “Interrupt Controller Modules” for more information.
Note that the UARTs can also be configured to automatically transfer data by using the DMA rather than interrupting the core. When there is data in the receiver FIFO or when the transmit holding register is empty, a DMA request can be issued. For more information on generating DMA requests, refer to
Section 26.4.6.1.2, “Setting up the UART to Request DMA Service ,” and
Chapter 16, “Enhanced Direct
Memory Access (eDMA).”
briefly describes the UART module signals.
NOTE
The terms ‘assertion’ and ‘negation’ are used to avoid confusion between active-low and active-high signals. ‘Asserted’ indicates that a signal is active, independent of the voltage level; ‘negated’ indicates that a signal is inactive.
Table 26-1. UART Module Signals
Signal Description
Transmitter Serial
Data Output
(U n TXD)
Receiver Serial
Data Input
(U n RXD)
Clear-to- Send
(U n CTS)
Request-to-Send
(U n RTS)
U n TXD is held high (mark condition) when the transmitter is disabled, idle, or operating in the local loop-back mode. Data is shifted out on U n TXD on the falling edge of the clock source, with the least significant bit (lsb) sent first.
Data received on U n RXD is sampled on the rising edge of the clock source, with the lsb received first.
This input can generate an interrupt on a change of state.
This output can be programmed to be negated or asserted automatically by either the receiver or the transmitter. When connected to a transmitter’s U n CTS, U n RTS can control serial data flow.
shows a signal configuration for a UART/RS-232 interface.
UART RS-232 Transceiver
U n RTS
U n CTS
U n TXD
U n RXD
DI2
DO2
DI1
DO1
Figure 26-2. UART/RS-232 Interface
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
26-3
UART Modules
26.3
Memory Map/Register Definition
This section contains a detailed description of each register and its specific function. Flowcharts in
Section 26.4.6, “Programming ,” describe basic UART module programming. The operation of the UART
module is controlled by writing control bytes into the appropriate registers.
for UART module registers.
NOTE
UART registers are accessible only as bytes.
NOTE
Interrupt can mean either an interrupt request asserted to the CPU or a DMA request.
Table 26-2. UART Module Memory Map
ISPBAR
Offset
UART0
UART1
UART2
0x00_0200
0x00_0240
0x00_0280
0x00_0204
0x00_0244
0x00_0284
0x00_0208
0x00_0248
0x00_0288
0x00_020C
0x00_024C
0x00_028C
0x00_0210
0x00_0250
0x00_0290
0x00_0214
0x00_0254
0x00_0294
0x00_0218
0x00_0258
0x00_0298
0x00_021C
0x00_025C
0x00_029C
Register
UART Mode Registers
1
(UMR1 n ), (UMR2 n )
UART Status Register (USR n ) 8
UART Clock Select Register
1
(UCSR n ) 8
UART Command Registers (UCR n )
UART Receive Buffers (URB n )
UART Transmit Buffers (UTB n )
UART Input Port Change Register (UIPCR n )
UART Auxiliary Control Register (UACR n )
UART Interrupt Status Register (UISR n )
UART Interrupt Mask Register (UIMR n )
UART Baud Rate Generator Register (UBG1 n )
UART Baud Rate Generator Register (UBG2 n )
Width
(bit)
8
8
8
8
8
8
8
8
8
8
Access Reset Value Section/Page
R/W
R
W
W
R
W
R
W
R
W
W
W
2
0x00
0x00
0x00
0x00
0xFF
0x00
See Section
0x00
0x00
0x00
0x00
0x00
26-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
UART Modules
Table 26-2. UART Module Memory Map (continued)
ISPBAR
Offset
Register
Width
(bit)
Access Reset Value Section/Page
UART0
UART1
UART2
0x00_0234
0x00_0274
0x00_02B4
UART Input Port Register (UIP n ) 8 R 0xFF
0x00_0238
0x00_0278
0x00_02B8
UART Output Port Bit Set Command Register (UOP1 n ) 8 W
0x00
0x00_023C
0x00_027C
0x00_02BC
UART Output Port Bit Reset Command Register (UOP0 n ) 8 W
0x00
1
2
UMR1 n , UMR2 n , and UCSR n should be changed only after the receiver/transmitter is issued a software reset command. That is, if operation is not disabled, undesirable results may occur.
Reading this register results in undesired effects and possible incorrect transmission or reception of characters. Register contents may also be changed.
26.3.1
UART Mode Registers 1 (UMR1
n
)
The UMR1 n registers control configuration. UMR1 n can be read or written when the mode register pointer points to it, at RESET or after a RESET MODE REGISTER POINTER command using UCR n [MISC]. After
UMR1 n is read or written, the pointer points to UMR2 n .
Address: 0x00_0200 (UMR10)
0x00_0240 (UMR11)
0x00_0280 (UMR12)
7 6 5 4 3
R
W
RXRTS
RXIRQ/
FFULL
ERR PM
Reset: 0 0 0 0 0
1
After UMR1 n is read or written, the pointer points to UMR2 n
Figure 26-3. UART Mode Registers 1 (UMR1 n )
2
PT
0
1
0
B/C
0
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
26-5
UART Modules
Table 26-3. UMR1 n Field Descriptions
Field Description
7
RXRTS
Receiver request-to-send. Allows the U n RTS output to control the U n CTS input of the transmitting device to prevent receiver overrun. If both the receiver and transmitter are incorrectly programmed for U n RTS control, U n RTS control is disabled for both. Transmitter RTS control is configured in UMR2 n [TXRTS].
0 The receiver has no effect on U n RTS.
1 When a valid start bit is received, U n RTS is negated if the UART's FIFO is full. U n RTS is reasserted when the
FIFO has an empty position available.
6
RXIRQ/
FFULL
Receiver interrupt select.
0 RXRDY is the source that generates interrupt or DMA requests.
1 FFULL is the source that generates interrupt or DMA requests.
5
ERR
Error mode. Configures the FIFO status bits, USR n [RB,FE,PE].
0 Character mode. The USR n values reflect the status of the character at the top of the FIFO. ERR must be 0 for correct A/D flag information when in multidrop mode.
1 Block mode. The USR n values are the logical OR of the status for all characters reaching the top of the FIFO since the last RESET ERROR STATUS
command for the UART was issued. See Section 26.3.5, “UART Command
4–3
PM
2
PT
Parity mode. Selects the parity or multidrop mode for the UART. The parity bit is added to the transmitted character, and the receiver performs a parity check on incoming data. The value of PM affects PT, as shown below.
Parity type. PM and PT together select parity type (PM = 0x) or determine whether a data or address character is transmitted (PM = 11).
PM
00
01
10
11
Parity Mode
With parity
Force parity
No parity
Multidrop mode
Parity Type (PT= 0) Parity Type (PT= 1)
Even parity
Low parity
Odd parity
High parity
N/A
Data character Address character
1–0
B/C
Bits per character. Selects the number of data bits per character to be sent. The values shown do not include start, parity, or stop bits.
00 5 bits
01 6 bits
10 7 bits
11 8 bits
26.3.2
UART Mode Register 2 (UMR2
n
)
The UMR2 n registers control UART module configuration. UMR2 n can be read or written when the mode register pointer points to it, which occurs after any access to UMR1 n . UMR2 n accesses do not update the pointer.
26-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
UART Modules
Address: 0x00_0200 (UMR20)
0x00_0240 (UMR21)
0x00_0280 (UMR22)
7 6 5 4 3
R
W
CM TXRTS TXCTS
Reset: 0 0 0 0 0
1
After UMR1 n is read or written, the pointer points to UMR2 n
Figure 26-5. UART Mode Register 2 (UMR2 n )
2
0
SB
1
0
0
0
Table 26-4. UMR2 n Field Descriptions
Field Description
7–6
CM
Channel mode. Selects a channel mode. Section 26.4.3, “Looping Modes ,” describes individual modes.
00 Normal
01 Automatic echo
10 Local loop-back
11 Remote loop-back
5
TXRTS
Transmitter ready-to-send. Controls negation of U n RTS to automatically terminate a message transmission.
Attempting to program a receiver and transmitter in the same UART for U n RTS control is not permitted and disables
U n RTS control for both.
0 The transmitter has no effect on U n RTS.
1 In applications where the transmitter is disabled after transmission completes, setting this bit automatically clears
UOP[RTS] one bit time after any characters in the transmitter shift and holding registers are completely sent, including the programmed number of stop bits.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
26-7
UART Modules
Table 26-4. UMR2 n Field Descriptions (continued)
Field Description
4
TXCTS
Transmitter clear-to-send. If both TXCTS and TXRTS are set, TXCTS controls the operation of the transmitter.
0 U n CTS has no effect on the transmitter.
1 Enables clear-to-send operation. The transmitter checks the state of U n CTS each time it is ready to send a character. If U n CTS is asserted, the character is sent; if it is deasserted, the signal U n TXD remains in the high state and transmission is delayed until U n CTS is asserted. Changes in U n CTS as a character is being sent do not affect its transmission.
3–0
SB
Stop-bit length control. Selects the length of the stop bit appended to the transmitted character. Stop-bit lengths of
9/16 to 2 bits are programmable for 6–8 bit characters. Lengths of 1-1/16 to 2 bits are programmable for 5-bit characters. In all cases, the receiver checks only for a high condition at the center of the first stop-bit position, that is, one bit time after the last data bit or after the parity bit, if parity is enabled. If an external 1x clock is used for the transmitter, clearing bit 3 selects one stop bit and setting bit 3 selects two stop bits for transmission.
SB
0000
0001
0010
0011
0100
0101
0110
0111
5 Bits 6–8 Bits
1.063
1.125
1.188
1.250
1.313
1.375
1.438
1.500
0.563
0.625
0.688
0.750
0.813
0.875
0.938
1.000
SB
1000
1001
1010
1011
1100
1101
1110
1111
5–8 Bits
1.563
1.625
1.688
1.750
1.813
1.875
1.938
2.000
26.3.3
UART Status Registers (USR
n
)
The USR n registers, shown in
, show the status of the transmitter, the receiver, and the FIFO.
Access: User read-only Address: 0x00_0204 (UCSR0)
0x00_0244 (UCSR1)
0x00_0284 (UCSR2)
7
RB R
W
Reset: 0
6
FE
5
PE
4
OE
3
TXEMP
2
TXRDY
0 0 0 0
Figure 26-6. UART Status Register (USR n )
0
1
FFULL
0
0
RXRDY
0
26-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
UART Modules
Table 26-5. USR n Field Descriptions
Field Description
7
RB
6
FE
5
PE
4
OE
Received break. The received break circuit detects breaks that originate in the middle of a received character.
However, a break in the middle of a character must persist until the end of the next detected character time.
0 No break was received.
1 An all-zero character of the programmed length was received without a stop bit. Only a single FIFO position is occupied when a break is received. Further entries to the FIFO are inhibited until U n RXD returns to the high state for at least one-half bit time, which is equal to two successive edges of the UART clock. RB is valid only when
RXRDY = 1.
Framing error.
0 No framing error occurred.
1 No stop bit was detected when the corresponding data character in the FIFO was received. The stop-bit check occurs in the middle of the first stop-bit position. FE is valid only when RXRDY = 1.
Parity error. Valid only if RXRDY = 1.
0 No parity error occurred.
1 If UMR1 n [PM] = 0 x (with parity or force parity), the corresponding character in the FIFO was received with incorrect parity. If UMR1 n [PM] = 11 (multidrop), PE stores the received address or data (A/D) bit. PE is valid only when RXRDY = 1.
Overrun error. Indicates whether an overrun occurs.
0 No overrun occurred.
1 One or more characters in the received data stream have been lost. OE is set upon receipt of a new character when the FIFO is full and a character is already in the shift register waiting for an empty FIFO position. When this occurs, the character in the receiver shift register and its break detect, framing error status, and parity error, if any, are lost. OE is cleared by the RESET ERROR STATUS command in UCR n .
3
TEMP
2
TXRDY
1
FFULL
Transmitter empty.
0 The transmit buffer is not empty. Either a character is being shifted out, or the transmitter is disabled. The transmitter is enabled/disabled by programming UCR n [TC].
1 The transmitter has underrun (both the transmitter holding register and transmitter shift registers are empty). This bit is set after transmission of the last stop bit of a character if there are no characters in the transmitter holding register awaiting transmission.
Transmitter ready.
0 The CPU loaded the transmitter holding register or the transmitter is disabled.
1 The transmitter holding register is empty and ready for a character. TXRDY is set when a character is sent to the transmitter shift register or when the transmitter is first enabled. If the transmitter is disabled, characters loaded into the transmitter holding register are not sent.
FIFO full.
0 The FIFO is not full but may hold up to two unread characters.
1 A character was received and the receiver FIFO is now full. Any characters received when the FIFO is full are lost.
0
RXRDY
Receiver ready.
0 The CPU has read the receive buffer and no characters remain in the FIFO after this read.
1 One or more characters were received and are waiting in the receive buffer FIFO.
26.3.4
UART Clock Select Registers (UCSR
n
)
The UCSRs select an external clock on the DTIN input (divided by 1 or 16) or a prescaled internal bus clock as the clocking source for the transmitter and receiver. See
Section 26.4.1, “Transmitter/Receiver
.” The transmitter and receiver can use different clock sources. To use the internal bus clock for both, set UCSR n to 0xDD.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
26-9
UART Modules
Field
7–4
RCS
Address: 0x00_0204 (UCSR0)
0x00_0244 (UCSR1)
0x00_0284 (UCSR2)
7 6
R
W
Reset: 0
RCS
5 4 3 2
0 0 0 0 0
Figure 26-7. UART Clock Select Register (UCSR n )
TCS
Access: User write-only
1
0
0
0
3–0
TCS
Table 26-6. UCSR n Field Descriptions
Description
Receiver clock select. Selects the clock source for the receiver.
1101 Prescaled internal bus clock (f sys/2
)
1110 DTIN divided by 16
1111 DTIN
Transmitter clock select. Selects the clock source for the transmitter.
1101 Prescaled internal bus clock (f sys/2
)
1110 DTIN divided by 16
1111 DTIN
26.3.5
UART Command Registers (UCR
n
)
The UCRs, shown in Figure 26-8 , supply commands to the UART. Only multiple commands that do not
conflict can be specified in a single write to a UCR n . For example,
RESET TRANSMITTER
and
ENABLE
TRANSMITTER
cannot be specified in one command.
Access: User write-only Address: 0x00_0208 (UCR0)
0x00_0248 (UCR1)
0x00_0288 (UCR2)
7 6
R
W
Reset:
0
0
5 4 3 2
0
MISC
0 0 0
TC
Figure 26-8. UART Command Register (UCR n )
0
1
0
RC
0
0
26-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
UART Modules
describes UCR n
fields and commands. Examples in Section 26.4.2, “Transmitter and Receiver
,” show how these commands are used.
Table 26-7. UCR n Field Descriptions
Field
7
6–4
MISC
Reserved, should be cleared.
MISC Field (this field selects a single command)
Description
Command Description
000 NO COMMAND —
001 RESET MODE
REGISTER POINTER
Causes the mode register pointer to point to UMR1 n .
010
011
111
RESET
RESET
TRANSMITTER
STOP
RECEIVER
BREAK
Immediately disables the receiver, clears USR n [FFULL,RXRDY], and reinitializes the receiver FIFO pointer. No other registers are altered. Because it places the receiver in a known state, use this command instead of RECEIVER DISABLE when reconfiguring the receiver.
Immediately disables the transmitter and clears USR n [TXEMP,TXRDY]. No other registers are altered. Because it places the transmitter in a known state, use this command instead of TRANSMITTER DISABLE when reconfiguring the transmitter.
Clears USR n [RB,FE,PE,OE]. Also used in block mode to clear all error bits after a data block is received.
100 RESET ERROR
STATUS
101 RESET BREAK –
CHANGE INTERRUPT
Clears the delta break bit, UISR n [DB].
110 START BREAK Forces U n TXD low. If the transmitter is empty, the break may be delayed up to one bit time. If the transmitter is active, the break starts when character transmission completes. The break is delayed until any character in the transmitter shift register is sent. Any character in the transmitter holding register is sent after the break. The transmitter must be enabled for the command to be accepted. This command ignores the state of U n CTS.
Causes U n TXD to go high (mark) within two bit times. Any characters in the transmit buffer are sent.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
26-11
UART Modules
Table 26-7. UCR n Field Descriptions (continued)
Field
3–2
TC
TC Field (This field selects a single command)
Description
Command Description
00 NO ACTION TAKEN Causes the transmitter to stay in its current mode: if the transmitter is enabled, it remains enabled; if the transmitter is disabled, it remains disabled.
01 TRANSMITTER
ENABLE
Enables operation of the UART’s transmitter. USR n [TXEMP,TXRDY] are set. If the transmitter is already enabled, this command has no effect.
10 TRANSMITTER
DISABLE
11 —
Terminates transmitter operation and clears USR transmitter becomes inactive. If the transmitter is already disabled, the command has no effect.
Reserved, do not use.
n [TXEMP,TXRDY]. If a character is being sent when the transmitter is disabled, transmission completes before the
1–0
RC
RC (This field selects a single command)
Command Description
00 NO ACTION TAKEN Causes the receiver to stay in its current mode. If the receiver is enabled, it remains enabled; if disabled, it remains disabled.
01 RECEIVER ENABLE If the UART module is not in multidrop mode (UMR1 n [PM]
≠
11), RECEIVER ENABLE enables the UART's receiver and forces it into search-for-start-bit state. If the receiver is already enabled, this command has no effect.
10 RECEIVER DISABLE Disables the receiver immediately. Any character being received is lost. The command does not affect receiver status bits or other control registers. If the
UART module is programmed for local loop-back or multidrop mode, the receiver operates even though this command is selected. If the receiver is already disabled, the command has no effect.
11 — Reserved, do not use.
26.3.6
UART Receive Buffers (URB
n
)
The receive buffers (shown in Figure 26-9
) contain one serial shift register and three receiver holding registers, which act as a FIFO. U n RXD is connected to the serial shift register. The CPU reads from the top of the FIFO while the receiver shifts and updates from the bottom when the shift register is full (see
). RB contains the character in the receiver.
26-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
UART Modules
Address: 0x00_020C (URB0)
0x00_024C (URB1)
0x00_028C (URB2)
7 6
R
W
Reset: 1
5 4
RB
3 2
1 1 1 1
Figure 26-9. UART Receive Buffer (URB n )
1
Access: User read-only
1
1
0
1
26.3.7
UART Transmit Buffers (UTB
n
)
The transmit buffers consist of the transmitter holding register and the transmitter shift register. The holding register accepts characters from the bus master if UART’s USR n [TXRDY] is set. A write to the transmit buffer clears USR n [TXRDY], inhibiting any more characters until the shift register can accept more data. When the shift register is empty, it checks if the holding register has a valid character to be sent
(TXRDY = 0). If there is a valid character, the shift register loads it and sets USR n [TXRDY] again. Writes to the transmit buffer when the UART’s TXRDY = 0 and the transmitter is disabled have no effect on the transmit buffer.
shows UTB n . TB contains the character in the transmit buffer.
Access: User write-only Address: 0x00_020C (UTB0)
0x00_024C (UTB1)
0x00_028C (UTB2)
7 6
R
W
Reset: 0
5 4 3 2
TB
0 0 0 0
Figure 26-10. UART Transmit Buffer (UTB n )
0
1
0
0
0
26.3.8
UART Input Port Change Registers (UIPCR
n
)
The UIPCRs, shown in Figure 26-11
, hold the current state and the change-of-state for U n CTS.
Address: 0x00_0210 (UIPCR0)
0x00_0250 (UIPCR1)
0x00_0290 (UIPCR2)
7
0 R
W
Reset: 0
6
0
5
0
4
COS
3
1
2
1
0 0 0 1 1
Figure 26-11. UART Input Port Change Register (UIPCR n )
Access: User read-only
1
1
1
0
CTS
U n CTS
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
26-13
UART Modules
Table 26-8. UIPCR n Field Descriptions
Field
7–5
4
COS
3–1
0
CTS
Description
Reserved
Change of state (high-to-low or low-to-high transition).
0 No change-of-state since the CPU last read UIPCR n . Reading UIPCR n clears UISR n [COS].
1 A change-of-state longer than 25–50
µ s occurred on the U n CTS input. UACR n can be programmed to generate an interrupt to the CPU when a change of state is detected.
Reserved
Current state of clear-to-send. Starting two serial clock periods after reset, CTS reflects the state of U n CTS. If
U n CTS is detected asserted at that time, COS is set, which initiates an interrupt if UACR n [IEC] is enabled.
0 The current state of the U n CTS input is asserted.
1 The current state of the U n CTS input is deasserted.
26.3.9
UART Auxiliary Control Register (UACR
n
)
The UACRs, shown in Figure 26-9 , control the input enable.
Address: 0x00_0210 (UACR0)
0x00_0250 (UACR1)
0x00_0290 (UACR2)
7 6
R
W
Reset:
0
0
0
0
5
0
0
4
0
0
3
0
0
2
0
0
Figure 26-12. UART Auxiliary Control Register (UACR n )
Access: User write-only
1
0
0
0
IEC
0
Field
7–1
0
IEC
Table 26-9. UACR n Field Descriptions
Description
Reserved, should be cleared.
Input enable control.
0 Setting the corresponding UIPCR n bit has no effect on UISR n [COS].
1 UISR n [COS] is set and an interrupt is generated when the UIPCR n [COS] is set by an external transition on the
U n CTS input (if UIMR n [COS] = 1).
26.3.10 UART Interrupt Status/Mask Registers (UISR
n
/UIMR
n
)
The UISRs, shown in Figure 26-13 , provide status for all potential interrupt sources. UISR
n contents are masked by UIMR n . If corresponding UISR n and UIMR n bits are set, the internal interrupt output is asserted. If a UIMR n bit is cleared, the state of the corresponding UISR n bit has no effect on the output.
The UISR n and UIMR n registers share the same space in memory. Reading this register provides the user with interrupt status, while writing controls the mask bits.
26-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
UART Modules
NOTE
True status is provided in the UISR n regardless of UIMR n settings. UISR n is cleared when the UART module is reset.
Address: 0x00_0214 (UISR0)
0x00_0254 (UISR1)
0x00_0294 (UISR2)
R
(UISR n )
W
(UIMR n )
Reset:
Access: User read/write
7
COS
COS
6
0
0
5
0
0
4
0
0
3
0
0
2
DB
DB
1
FFULL/
RXRDY
FFULL/
RXRDY
0 0 0 0 0 0 0
Figure 26-13. UART Interrupt Status/Mask Registers (UISR n /UIMR n )
0
TXRDY
TXRDY
0
Table 26-10. UISR n /UIMR n Field Descriptions
Field Description
7
COS
Change-of-state.
0 UIPCR n [COS] is not selected.
1 Change-of-state occurred on U n CTS and was programmed in UACR n [IEC] to cause an interrupt.
Reserved, should be cleared.
6–3
2
DB
Delta break.
0 No new break-change condition to report.
Section 26.3.5, “UART Command Registers (UCRn) ,” describes the
RESET BREAK CHANGE INTERRUPT command.
1 The receiver detected the beginning or end of a received break.
1
FFULL/
RXRDY
Status of FIFO or receiver, depending on UMR1[FFULL/RXRDY] bit. Duplicate of USR n [FIFO] & USR n [RXRDY]
UMR1 n [FFULL/RXRDY]
UIMR n
[FFULL/RXRDY]
UISR n
[FFULL/RXRDY]
0 (RXRDY) 1 (FIFO)
0
1
0
1
0
0
1
1
Receiver not ready
Receiver not ready
Receiver is ready,
Do not interrupt
Receiver is ready, interrupt
FIFO not full
FIFO not full
FIFO is full,
Do not interrupt
FIFO is full, interrupt
0
TXRDY
Transmitter ready. This bit is the duplication of USR n [TXRDY].
0 The transmitter holding register was loaded by the CPU or the transmitter is disabled. Characters loaded into the transmitter holding register when TXRDY = 0 are not sent.
1 The transmitter holding register is empty and ready to be loaded with a character.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
26-15
UART Modules
26.3.11 UART Baud Rate Generator Registers (UBG1
n
/UBG2
n
)
The UBG1 n registers hold the MSB, and the UBG2 n registers hold the LSB of the preload value. UBG1 n and UBG2 n concatenate to provide a divider to the internal bus clock for transmitter/receiver operation,
as described in Section 26.4.1.2.1, “Internal Bus Clock Baud Rates .”
Address: 0x00_0218 (UBG10)
0x00_0258 (UBG11)
0x00_0298 (UBG12)
7 6
R
W
Reset: 0 0
5
0
4 3
Divider MSB
0 0
2
0
Figure 26-14. UART Baud Rate Generator Register (UBG1 n )
Access: User write-only
1
0
0
0
Address: 0x00_021C (UBG20)
0x00_025C (UBG21)
0x00_029C (UBG22)
7 6
R
W
Reset:
5 4 3 2
Access: User write-only
1
0 0 0 0
Divider LSB
0 0 0
Figure 26-15. UART Baud Rate Generator Register (UBG2 n )
NOTE
The minimum value that can be loaded on the concatenation of UBG1 n with
UBG2 n is 0x0002. The UBG2 n reset value of 0x00 is invalid and must be written to before the UART transmitter or receiver are enabled. Both
UBG1 n and UBG2 n are write-only and cannot be read by the CPU.
0
0
26.3.12 UART Input Port Register (UIP
n
)
The UIP n registers, shown in
, show the current state of the U n CTS input.
26-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
UART Modules
Field
7–1
0
CTS
Address: 0x00_0234 (UIP0)
0x00_0274 (UIP1)
0x00_02B4 (UIP2)
7
1 R
W
Reset: 1
6
1
5
1
4
1
3
1
1 1 1 1
Figure 26-17. UART Input Port Register (UIP n )
1
2
1
Access: User read-only
1
1
1
0
CTS
1
Table 26-12. UIP n Field Descriptions
Description
Reserved
Current state of clear-to-send. The U n CTS value is latched and reflects the state of the input pin when UIP n is read.
Note: This bit has the same function and value as UIPCR n [RTS].
0 The current state of the U n CTS input is logic 0.
1 The current state of the U n CTS input is logic 1.
26.3.13 UART Output Port Command Registers (UOP1
n
/UOP0
n
)
The U n RTS output can be asserted by writing a 1 to UOP1 n [RTS] and negated by writing a 1 to
UOP0 n
Address: 0x00_0238 (UOP10)
0x00_023C (UOP00)
0x00_0278 (UOP11)
0x00_027C (UOP01)
0x00_02B8 (UOP12)
0x00_02BC (UOP02)
7 6
R
W
Reset:
0
0
0
0
5
0
0
4
0
0
3
0
0
2
0
0
Access: User write-only
1
0
0
Figure 26-18. UART Output Port Command Registers (UOP1 n /UOP0 n )
0
RTS
0
Table 26-13. UOP1 n /UOP0 n Field Descriptions
Field
7–1
0
RTS
Description
Reserved, should be cleared.
Output port output. Controls assertion (UOP1)/negation (UOP0) of U n RTS output.
0 Not affected.
1 Asserts U n RTS in UOP1. Negates U n RTS in UOP0.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
26-17
UART Modules
26.4
Functional Description
This section describes operation of the clock source generator, transmitter, and receiver.
26.4.1
Transmitter/Receiver Clock Source
The internal bus clock serves as the basic timing reference for the clock source generator logic, which consists of a clock generator and a programmable 16-bit divider dedicated to each UART. The clock generator might not produce standard baud rates if the internal bus clock is used, so the user must enable the 16-bit divider.
26.4.1.1
Programmable Divider
As Figure 26-19 shows, the UART
n transmitter and receiver can use the following clock sources:
• An external clock signal on the DT n IN pin.
When not divided, DT
n
IN provides a synchronous clock; when divided by 16, it is asynchronous.
• The internal bus clock supplies an asynchronous clock source that is divided by 32 and then divided by the 16-bit value programmed in UBG1 n and UBG2 n
. See Section 26.3.11, “UART Baud Rate
Generator Registers (UBG1n/UBG2n) .”
The choice of DTIN or internal bus clock is programmed in the UCSR.
DT n OUT
DT n IN
On-Chip
Timer Module
U n TXD
Tx Buffer
UART
Clocking sources programmed in UCSR
÷
1
TIN
Tx
÷
16
TIN
Rx
Clock
Generator
16-Bit
Divider
÷
32
U n RXD Rx Buffer
Figure 26-19. Clocking Source Diagram
NOTE
If DT n IN is a clocking source for either the timer or UART, that timer module cannot use DT n IN for timer input capture.
Internal
Bus Clock f sys
26-18
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
UART Modules
26.4.1.2
Calculating Baud Rates
The following sections describe how to calculate baud rates.
26.4.1.2.1
Internal Bus Clock Baud Rates
When the internal bus clock is the UART clocking source, it goes through a divide-by-32 prescaler and then passes through the 16-bit divider of the concatenated UBG1 n and UBG2 n registers. The baud-rate calculation is as follows:
Baudrate
=
[ f
-----------------------------------
32 x divider
]
Using a 66 MHz internal bus clock and letting baud rate = 9600, then
Eqn. 26-1
Divider =
[
66
MHz
32 x 9600
]
=
215 ( decimal
)
=
0x00D6 ( hexadecimal
)
Therefore, UBG1 n = 0x00 and UBG2 n = 0xD6.
Eqn. 26-2
26.4.1.2.2
External Clock
An external source clock (DT n IN) passes through a divide-by-1 or 16 prescaler. If f extc frequency, then the baud rate can be described with this equation:
is the external clock
Baudrate = f
(16 or 1)
Eqn. 26-3
26.4.2
Transmitter and Receiver Operating Modes
is a functional block diagram of the transmitter and receiver showing the command and operating registers, which are described generally in the following sections. For detailed descriptions, refer
to Section 26.3, “Memory Map/Register Definition
.”
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
26-19
UART Modules
UART n
UART Command Register (UCR n ) W
UART Mode Register 1 (UMR1 n )
UART Mode Register 2 (UMR2 n )
R/W
R/W
UART Status Register (USR n )
R
UART
Transmit Buffer
(UTB n )
(2 Registers)
Transmitter Holding Register
Transmitter Shift Register
W
UART Receive
Buffer (URB n )
(4 Registers)
Receiver Holding Register 1
Receiver Holding Register 2
Receiver Holding Register 3
Receiver Shift Register
R
FIFO
Figure 26-20. Transmitter and Receiver Functional Diagram
U n TXD
External
Interface
U n RXD
26.4.2.1
Transmitter
The transmitter is enabled through the UART command register (UCR n ). When it is ready to accept a character, the UART sets USR n [TXRDY]. The transmitter converts parallel data from the CPU to a serial bit stream on U n TXD. It automatically sends a start bit followed by the programmed number of data bits, an optional parity bit, and the programmed number of stop bits. The lsb is sent first. Data is shifted from the transmitter output on the falling edge of the clock source.
After the stop bits are sent, if no new character is in the transmitter holding register, the U n TXD output remains high (mark condition) and the transmitter empty bit (USR n [TXEMP]) is set. Transmission resumes and TXEMP is cleared when the CPU loads a new character into the UART transmit buffer
(UTB n ). If the transmitter receives a disable command, it continues until any character in the transmitter shift register is completely sent.
If the transmitter is reset through a software command, operation stops immediately (see
“UART Command Registers (UCRn)
”). The transmitter is reenabled through the UCR n to resume operation after a disable or software reset.
If the clear-to-send operation is enabled, U n CTS must be asserted for the character to be transmitted. If
U n CTS is negated in the middle of a transmission, the character in the shift register is sent and U n TXD remains in mark state until U n CTS is reasserted. If the transmitter is forced to send a continuous low condition by issuing a SEND BREAK command, the transmitter ignores the state of U n CTS.
26-20
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
UART Modules
If the transmitter is programmed to automatically negate U n RTS when a message transmission completes,
U n RTS must be asserted manually before a message is sent. In applications in which the transmitter is disabled after transmission is complete and U n RTS is appropriately programmed, U n RTS is negated one bit time after the character in the shift register is completely transmitted. The transmitter must be manually reenabled by reasserting U n RTS before the next message is to be sent.
shows the functional timing information for the transmitter.
C1 in transmission
U n TXD C1
1
C2 C3 Break C4 C6
Transmitter
Enabled
USR n [TXRDY] internal module select
U n CTS
3
W
2
C1
1
W
C2
W W
C3 Start break
W W W
C4 Stop break
C5 not transmitted
W
C6
U n RTS
4 Manually asserted by BIT SET command
1
C n = transmit characters
2
W = write
3
UMR2 n [TXCTS] = 1
4
UMR2 n [TXRTS] = 1
Figure 26-21. Transmitter Timing Diagram
Manually asserted
26.4.2.2
Receiver
The receiver is enabled through its UCR n
, as described in Section 26.3.5, “UART Command Registers
When the receiver detects a high-to-low (mark-to-space) transition of the start bit on U n RXD, the state of
U n RXD is sampled eight times on the edge of the bit time clock starting one-half clock after the transition
(asynchronous operation) or at the next rising edge of the bit time clock (synchronous operation). If
U n RXD is sampled high, the start bit is invalid and the search for the valid start bit begins again.
If U n RXD is still low, a valid start bit is assumed and the receiver continues sampling the input at one-bit time intervals, at the theoretical center of the bit, until the proper number of data bits and parity, if any, is assembled and one stop bit is detected. Data on the U n RXD input is sampled on the rising edge of the
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
26-21
UART Modules programmed clock source. The lsb is received first. The data is then transferred to a receiver holding register and USR n [RXRDY] is set. If the character is less than 8 bits, the most significant unused bits in the receiver holding register are cleared.
After the stop bit is detected, the receiver immediately looks for the next start bit. However, if a non-zero character is received without a stop bit (framing error) and U n RXD remains low for one-half of the bit period after the stop bit is sampled, the receiver operates as if a new start bit were detected. Parity error, framing error, overrun error, and received break conditions set the respective PE, FE, OE, and RB error and break flags in the USR n at the received character boundary. They are valid only if USR n [RXRDY] is set.
If a break condition is detected (U n RXD is low for the entire character including the stop bit), a character of all zeros is loaded into the receiver holding register and USR n [RB,RXRDY] are set. U n RXD must return to a high condition for at least one-half bit time before a search for the next start bit begins.
The receiver detects the beginning of a break in the middle of a character if the break persists through the next character time. If the break begins in the middle of a character, the receiver places the damaged character in the Rx FIFO and sets the corresponding USR n error bits and USR n [RXRDY]. Then, if the break lasts until the next character time, the receiver places an all-zero character into the Rx FIFO and sets
USR n [RB,RXRDY].
shows receiver functional timing.
U n TXD C1 C2 C3 C4 C5 C6 C7 C8
C6, C7, and C8 will be lost
Receiver
Enabled
USR n [RXRDY]
USR n [FFULL] internal module select Status
Data
(C1)
Overrun
USR n [OE]
U n RTS
1
Manually asserted first time, automatically negated if overrun occurs
UOP0[RTS] = 1
1
UMR2 n [TXRTS] = 1
C5 will be lost
Figure 26-22. Receiver Timing
Status
Data
(C2)
Status
Data
Status
Data
(C3) (C4)
Reset by command
Automatically asserted when ready to receive
26-22
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
UART Modules
26.4.2.3
FIFO
The FIFO is used in the UART’s receive buffer logic. The FIFO consists of three receiver holding registers.
The receive buffer consists of the FIFO and a receiver shift register connected to the U n RXD (see
). Data is assembled in the receiver shift register and loaded into the top empty receiver holding register position of the FIFO. Thus, data flowing from the receiver to the CPU is quadruple-buffered.
In addition to the data byte, three status bits—parity error (PE), framing error (FE), and received break
(RB)—are appended to each data character in the FIFO; overrun error (OE) is not appended. By programming the ERR bit in the UART’s mode register (UMR1 n ), status is provided in character or block modes.
USR n [RXRDY] is set when at least one character is available to be read by the CPU. A read of the receive buffer produces an output of data from the top of the FIFO. After the read cycle, the data at the top of the
FIFO and its associated status bits are popped and the receiver shift register can add new data at the bottom of the FIFO. The FIFO-full status bit (FFULL) is set if all three positions are filled with data. Either the
RXRDY or FFULL bit can be selected to cause an interrupt and either TXRDY or RXRDY can be used to generate a DMA request.
The two error modes are selected by UMR1 n [ERR]:
• In character mode (UMR1 n [ERR] = 0, status is given in the USR n for the character at the top of the FIFO.
• In block mode, the USR n shows a logical OR of all characters reaching the top of the FIFO since the last RESET ERROR STATUS command. Status is updated as characters reach the top of the FIFO.
Block mode offers a data-reception speed advantage where the software overhead of error-checking each character cannot be tolerated. However, errors are not detected until the check is performed at the end of an entire message—the faulting character is not identified.
In either mode, reading the USR n does not affect the FIFO. The FIFO is popped only when the receive buffer is read. The USR n should be read before reading the receive buffer. If all three receiver holding registers are full, a new character is held in the receiver shift register until space is available. However, if a second new character is received, the contents of the character in the receiver shift register is lost, the
FIFOs are unaffected, and USR n [OE] is set when the receiver detects the start bit of the new overrunning character.
To support flow control, the receiver can be programmed to automatically negate and assert U n RTS, in which case the receiver automatically negates U n RTS when a valid start bit is detected and the FIFO is full. The receiver asserts U n RTS when a FIFO position becomes available; therefore, overrun errors can be prevented by connecting U n RTS to the U n CTS input of the transmitting device.
NOTE
The receiver can still read characters in the FIFO if the receiver is disabled.
If the receiver is reset, the FIFO, U n RTS control, all receiver status bits, and interrupts, and DMA requests are reset. No more characters are received until the receiver is reenabled.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
26-23
UART Modules
26.4.3
Looping Modes
are useful for local and remote system diagnostic functions. The modes are described in the following
paragraphs and in Section 26.3, “Memory Map/Register Definition
.”
The UART’s transmitter and receiver should be disabled when switching between modes. The selected mode is activated immediately upon mode selection, regardless of whether a character is being received or transmitted.
26.4.3.1
Automatic Echo Mode
In automatic echo mode, shown in
, the UART automatically resends received data bit by bit.
The local CPU-to-receiver communication continues normally, but the CPU-to-transmitter link is disabled. In this mode, received data is clocked on the receiver clock and re-sent on U n TXD. The receiver must be enabled, but the transmitter need not be.
Rx U n RXD Input
CPU
Disabled
Tx
Disabled
U n TXD Output
Figure 26-23. Automatic Echo
Because the transmitter is inactive, USR n [TXEMP,TXRDY] are inactive and data is sent as it is received.
Received parity is checked but not recalculated for transmission. Character framing is also checked, but stop bits are sent as they are received. A received break is echoed as received until the next valid start bit is detected.
26.4.3.2
Local Loop-back Mode
shows how U n TXD and U n RXD are internally connected in local loop-back mode. This mode is for testing the operation of a UART by sending data to the transmitter and checking data assembled by the receiver to ensure proper operations.
Rx
Disabled
U n RXD Input
CPU
Tx
Disabled
U n TXD Output
Figure 26-24. Local Loop-back
Features of this local loop-back mode are as follows:
• Transmitter and CPU-to-receiver communications continue normally in this mode.
• U n RXD input data is ignored.
• U n TXD is held marking.
• The receiver is clocked by the transmitter clock. The transmitter must be enabled, but the receiver need not be.
26-24
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
UART Modules
26.4.3.3
Remote Loop-back Mode
In remote loop-back mode, shown in Figure 26-25
, the UART automatically transmits received data bit by bit on the U n TXD output. The local CPU-to-transmitter link is disabled. This mode is useful in testing receiver and transmitter operation of a remote UART. For this mode, the transmitter uses the receiver clock.
Because the receiver is not active, received data cannot be read by the CPU and all status conditions are inactive. Received parity is not checked and is not recalculated for transmission. Stop bits are sent as they are received. A received break is echoed as received until the next valid start bit is detected.
Disabled
CPU
Disabled
Rx
Disabled
Tx
Disabled
Figure 26-25. Remote Loop-back
U n RXD Input
U n TXD Input
26.4.4
Multidrop Mode
Setting UMR1 n [PM] programs the UART to operate in a wake-up mode for multidrop or multiprocessor applications. In this mode, a master can transmit an address character followed by a block of data characters targeted for one of up to 256 slave stations.
Although slave stations have their receivers disabled, they continuously monitor the master’s data stream.
When the master sends an address character, the slave receiver notifies its respective CPU by setting
USR n [RXRDY] and generating an interrupt (if programmed to do so). Each slave station CPU then compares the received address to its station address and enables its receiver if it wishes to receive the subsequent data characters or block of data from the master station. Unaddressed slave stations continue monitoring the data stream. Data fields in the data stream are separated by an address character. After a slave receives a block of data, its CPU disables the receiver and repeats the process. Functional timing
information for multidrop mode is shown in Figure 26-26 .
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
26-25
UART Modules
U n TXD
Transmitter
Enabled
A/D
ADD1 1
Master Station
A/D
C0
A/D
ADD2 1
USR n [TXRDY] internal module select
UMR1 n [PM] = 11
UMR1 n [PT] = 1
ADD 1 C0
UMR1 n [PT] = 0
ADD 2
UMR1 n [PT] = 2
Peripheral Station
A/D A/D A/D
U n RXD 0 ADD1 1 C0
Receiver
Enabled
A/D
ADD2 1
A/D
0
USR n [RXRDY] internal module select
UMR1 n [PM] = 11
UMR1 n [PM] = 11
ADD 1 Status Data
(C0)
Figure 26-26. Multidrop Mode Timing Diagram
Status Data
(ADD 2)
A character sent from the master station consists of a start bit, a programmed number of data bits, an address/data (A/D) bit flag, and a programmed number of stop bits. A/D = 1 indicates an address character;
A/D = 0 indicates a data character. The polarity of A/D is selected through UMR1 n [PT]. UMR1 n should be programmed before enabling the transmitter and loading the corresponding data bits into the transmit buffer.
In multidrop mode, the receiver continuously monitors the received data stream, regardless of whether it is enabled or disabled. If the receiver is disabled, it sets the RXRDY bit and loads the character into the receiver holding register FIFO provided the received A/D bit is a one (address tag). The character is discarded if the received A/D bit is zero (data tag). If the receiver is enabled, all received characters are transferred to the CPU through the receiver holding register during read operations.
In either case, the data bits are loaded into the data portion of the FIFO while the A/D bit is loaded into the status portion of the FIFO normally used for a parity error (USR n [PE]).
Framing error, overrun error, and break detection operate normally. The A/D bit takes the place of the parity bit; therefore, parity is neither calculated nor checked. Messages in this mode may still contain error detection and correction information. If 8-bit characters are not required, one way to provide error detection is to use software to calculate parity and append it to the 5-, 6-, or 7-bit character.
26-26
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
UART Modules
26.4.5
Bus Operation
This section describes bus operation during read, write, and interrupt acknowledge cycles to the UART module.
26.4.5.1
Read Cycles
The UART module responds to reads with byte data. Reserved registers return zeros.
26.4.5.2
Write Cycles
The UART module accepts write data as bytes only. Write cycles to read-only or reserved registers complete normally without exception processing, but data is ignored.
26.4.6
Programming
The software flowchart,
, consists of the following:
• UART module initialization—These routines consist of SINIT and CHCHK (See Sheet 1 p. 26-30 and Sheet 2 p. 26-31). Before SINIT is called at system initialization, the calling routine allocates
2 words on the system FIFO. On return to the calling routine, SINIT passes UART status data on the FIFO. If SINIT finds no errors, the transmitter and receiver are enabled. SINIT calls CHCHK to perform the checks. When called, SINIT places the UART in local loop-back mode and checks for the following errors:
— Transmitter never ready
— Receiver never ready
— Parity error
— Incorrect character received
• I/O driver routine—This routine (See Sheet 4 p. 26-33 and Sheet 5 p. 26-34) consists of INCH, the terminal input character routine which gets a character from the receiver, and OUTCH, which sends a character to the transmitter.
• Interrupt handling—This consists of SIRQ (See Sheet 4 p. 26-33), which is executed after the
UART module generates an interrupt caused by a change-in-break (beginning of a break). SIRQ then clears the interrupt source, waits for the next change-in-break interrupt (end of break), clears the interrupt source again, then returns from exception processing to the system monitor.
26.4.6.1
Interrupt and DMA Request Initialization
26.4.6.1.1
Setting up the UART to Generate Core Interrupts
The list below gives the steps needed to properly initialize the UART to generate an interrupt request to the core. See Section 14.2.9.1, “Interrupt Sources,” for details on interrupt assignments for the UART modules.
1. Initialize the appropriate ICR x register in the interrupt controller.
2. Unmask appropriate bits in IMR in the interrupt controller.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
26-27
UART Modules
3. Unmask appropriate bits in the core’s status register (SR) to enable interrupts.
4. If TXRDY or RXRDY are being used to generate interrupt requests, then verify that DMAREQC
(in the SCM) does not also assign the UART’s TXRDY and RXRDY into DMA channels.
5. Initialize interrupts in the UART, see
.
Table 26-14. UART Interrupts
Register
UMR1 n
UIMR n
UIMR n
UIMR n
UIMR n
Bit
2
1
6
7
0
Interrupt
RxIRQ
Change of State (COS)
Delta Break
RxFIFO Full
TXRDY
26.4.6.1.2
Setting up the UART to Request DMA Service
The UART is capable of generating two different DMA request signals: transmit and receive.
The transmit DMA request signal is asserted when the TXRDY (transmitter ready) in the UART interrupt status register (UISR n [TXRDY]) is set. When the transmit DMA request signal is asserted, the DMA can initiate a data copy, reading the next character to be transmitted from memory and writing it into the UART transmit buffer (UTB n ). This would allow the DMA channel to stream data from memory to the UART for transmission without processor intervention. Once the entire message has been moved into the UART, the DMA would typically generate an end-of-data-transfer interrupt request to the CPU. The resulting interrupt service routine (ISR) could query the UART programming model to determine the end-of-transmission status.
Similarly, the receive DMA request signal is asserted when the FIFO full or receive ready
(FFULL/RXRDY) flag in the interrupt status register (UISR n [FFULL/RXRDY]) is set. When the receive
DMA request signal is asserted, the DMA can initiate a data move, reading the appropriate characters from the UART receive buffer (URB n ) and storing them in memory. This allows the DMA channel to stream data from the UART receive buffer into memory without processor intervention. Once the entire message has been moved from the UART, the DMA would typically generate an end-of-data-transfer interrupt request to the CPU. The resulting interrupt service routine (ISR) should query the UART programming model to determine the end-of-transmission status. In typical applications, the receive DMA request should be configured to use RXRDY directly (and not FFULL) to remove any complications related to retrieving the final characters from the FIFO buffer.
The implementation described in this section allows independent DMA processing of transmit and receive data while still supporting interrupt notification to the processor for CTS change-of-state and delta break error handling.
26-28
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
UART Modules
To configure the UART for DMA requests:
1. Initialize the DMAREQC in the SCM to map the desired UART DMA requests to the desired DMA channels. For example; setting DMAREQC[7:4] to 1000 maps UART0 receive DMA requests to
DMA channel 1; setting DMAREQC[11:8] to 1101 maps UART1 transmit DMA requests to DMA channel 2; and so on. It is possible to independently map transmit based and receive based UART
DMA requests in the DMAREQC.
2. Disable interrupts using the UIMR register. The appropriate UIMR bits must be cleared so that interrupt requests are disabled for those conditions for which a DMA request is desired. For example; to generate transmit DMA requests from UART1, then UIMR1[TXRDY] should be cleared. This will prevent TXRDY from generating an interrupt request while a transmit DMA request is generated.
3. Configure the GPACR and appropriate PACR registers located in the SCM for DMA access to
IPSBAR space.
4. Initialize the DMA channel. The DMA should be configured for cycle steal mode and a source and destination size of one byte. This will cause a single byte to be transferred for each UART DMA request.
shows the DMA requests.
Table 26-15. UART DMA Requests
Register
UISR n
UISR n
Bit
1
0
Receive DMA request
Transmit DMA request
DMA Request
26.4.6.2
UART Module Initialization Sequence
shows the UART module initialization sequence.
Table 26-16. UART Module Initialization Sequence
Register Setting
UCR n
UIMR n
UACR n
Reset the receiver and transmitter.
Reset the mode pointer (MISC[2–0] = 0b001).
Enable the desired interrupt sources.
Initialize the input enable control (IEC bit).
UCSR n Select the receiver and transmitter clock. Use timer as source if required.
UMR1 n If preferred, program operation of receiver ready-to-send (RXRTS bit).
Select receiver-ready or FIFO-full notification (RXRDY/FFULL bit).
Select character or block error mode (ERR bit).
Select parity mode and type (PM and PT bits).
Select number of bits per character (B/Cx bits).
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
26-29
UART Modules
Table 26-16. UART Module Initialization Sequence (continued)
Register Setting
UMR2 n Select the mode of operation (CMx bits).
If preferred, program operation of transmitter ready-to-send (TXRTS).
If preferred, program operation of clear-to-send (TXCTS bit).
Select stop-bit length (SBx bits).
UCR n Enable transmitter and/or receiver.
Serial Module
SINIT
Initiate:
Channel
Interrupts
CHK1
Call CHCHK
Save Channel
Status
Enable
Any
Errors?
N
Enable Receiver
Y
Assert
Request To Send
SINITR
Return
Figure 26-27. UART Mode Programming Flowchart (Sheet 1 of 5)
26-30
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
UART Modules
CHCHK
CHCHK
Place Channel In
Local Loopback
Mode
Enable
Transmitter Clear
Status Word
TxCHK
Is
Transmitter
Ready?
SNDCHR Y
Send Character
To Transmitter
RxCHK
Has
Character Been
Received?
Y
A
N
N
N
Waited
Too Long?
N
Waited
Too Long?
Y Set Transmitter-
Never-ready Flag
Y
Set Receiver-
Never-ready Flag
Figure 26-27. UART Mode Programming Flowchart (Sheet 2 of 5)
B
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
26-31
UART Modules
A
FRCHK
Have
Framing Error?
N
Set Framing
Error Flag
PRCHK
Have
Parity Error?
Y
Set Parity
Error Flag
CHRCHK
Get Character
From Receiver
N
Same As
Transmitted
Character?
N
Set Incorrect
Character Flag
Y
RSTCHN
B
Disable
Transmitter
Restore
To Original Mode
Return
B
Figure 26-27. UART Mode Programming Flowchart (Sheet 3 of 5)
26-32
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
ABRKI
SIRQ
Was
IRQ Caused
By Beginning
Of A Break?
Y
Clear Change-in-
Break Status Bit
ABRKI1
Has
End-of-break
IRQ Arrived
Yet?
Y
Clear Change-in-
Break Status Bit
N
N
Remove Break
Character From
Receiver FIFO
INCH
Does
Channel A
Receiver Have A
Character?
Y
Place Character
In D0
N
Return
Replace Return
Address On System
Stack And Monitor
Warm Start Address
SIRQR
RTE
Figure 26-27. UART Mode Programming Flowchart (Sheet 4 of 5)
UART Modules
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
26-33
UART Modules
OUTCH
Is
Transmitter
Ready?
Y
Send Character
To Transmitter
N
Return
Figure 26-27. UART Mode Programming Flowchart (Sheet 5 of 5)
26-34
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
I
Chapter 27
2
C Interface
27.1
Introduction
This chapter describes the I
2
C module, clock synchronization, and I
2 also provides extensive programming examples.
C programming model registers. It
27.2
Overview
I
2
C is a two-wire, bidirectional serial bus that provides a simple, efficient method of data exchange, minimizing the interconnection between devices. This bus is suitable for applications that require occasional communication between many devices over a short distance. The flexible I
2
C bus allows additional devices to be connected to the bus for expansion and system development.
The interface is designed to operate up to 100 Kbps with maximum bus loading and timing. The device is capable of operating at higher baud rates, up to a maximum of the internal bus clock divided by 20, with reduced bus loading. The maximum communication length and the number of devices that can be connected are limited by a maximum bus capacitance of 400 pF.
The I
2
C system is a true multiple-master bus; it uses arbitration and collision detection to prevent data corruption in the event that multiple devices attempt to control the bus simultaneously. This feature supports complex applications with multiprocessor control and can be used for rapid testing and alignment of end products through external connections to an assembly-line computer.
The I
2
NOTE
C module is designed to be compatible with the Philips I
2
C bus protocol. For information on system configuration, protocol, and restrictions, see The I
2
C Bus Specification, Version 2.1
.
NOTE
The GPIO module must be configured to enable the peripheral function of the appropriate pins (refer to Chapter 13, “General Purpose I/O Module” ) prior to configuring the I
2
C module.
27.3
Features
The I
2
C module has the following key features:
• Compatibility with I
2
C bus standard version 2.1
• Support for 3.3-V tolerant devices
• Multiple-master operation
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
27-1
I
2
C Interface
• Software-programmable for one of 50 different serial clock frequencies
• Software-selectable acknowledge bit
• Interrupt-driven, byte-by-byte data transfer
• Arbitration-lost interrupt with automatic mode switching from master to slave
• Calling address identification interrupt
• START and STOP signal generation/detection
• Repeated START signal generation
• Acknowledge bit generation/detection
• Bus-busy detection
is a block diagram of the I
2
C module.
IRQ
Internal Bus
Address Data
Registers and ColdFire Interface
Address Decode Data MUX
I
2
C Frequency
Divider Register
(IFDR)
Clock
Control
I
2
C Control
Register
(I2CR)
Start, Stop, and
Arbitration
Control
I
2
C Status
Register
(I2SR)
I
2
C Data
I/O Register
(I2DR)
I
2
C Address
Register
(IADR)
In/Out
Data
Shift
Register
Input
Sync
Address
Compare
I2C_SCL I2C_SDA
Figure 27-1. I
2
C Module Block Diagram
shows the I
:
2
C registers, which are described in Section 27.5, “Memory Map/Register
• I
2
C address register (I2ADR)
• I
2
C frequency divider register (I2FDR)
27-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
I
2
C Interface
• I
2
C control register (I2CR)
• I
2
C status register (I2SR)
• I
2
C data I/O register (I2DR)
27.4
I
2
C System Configuration
The I
2
C module uses a serial data line (I2C_SDA) and a serial clock line (I2C_SCL) for data transfer. For
I
2
C compliance, all devices connected to these two signals must have open drain or open collector outputs.
The logic AND function is exercised on both lines with external pull-up resistors.
Out of reset, the I
2
C default state is as a slave receiver. Thus, when not programmed to be a master or responding to a slave transmit address, the I
2
C module should return to the default slave receiver state. See
Section 27.6.1, “Initialization Sequence,”
for exceptions.
Normally, a standard communication is composed of four parts: START signal, slave address transmission, data transfer, and STOP signal. These are discussed in the following sections.
27.4.1
START Signal
When no other device is bus master (both I2C_SCL and I2C_SDA lines are at logic high), a device can initiate communication by sending a START signal (see A in
). A START signal is defined as a high-to-low transition of I2C_SDA while I2C_SCL is high. This signal denotes the beginning of a data transfer (each data transfer can be several bytes long) and awakens all slaves.
I2C_SCL msb
1 2 3 4 5 6
Interrupt bit set
(Byte complete)
I2C_SCL held low while
Interrupt is serviced lsb
7 8 msb
2 3 4 5 6 lsb
7 8
I2C_SDA AD7 AD6 AD5 AD4 AD3 AD2 AD1 R/W
A
START
Signal
XXX D7 D6 D5 D4 D3 D2 D1 D0
Calling Address
B
E
Data Byte
R/W ACK
C
Bit
D
Figure 27-2. I
2
C Standard Communication Protocol
No
ACK
Bit
STOP
Signal
F
27.4.2
Slave Address Transmission
The master sends the slave address in the first byte after the START signal (B). After the seven-bit calling address, it sends the R/W bit (C), which tells the slave data transfer direction (0 = write transfer, 1 = read transfer).
Each slave must have a unique address. An I
2
C master must not transmit its own slave address; it cannot be master and slave at the same time.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
27-3
I
2
C Interface
The slave whose address matches that sent by the master pulls I2C_SDA low at the ninth serial clock (D) to return an acknowledge bit.
27.4.3
Data Transfer
When successful slave addressing is achieved, the data transfer can proceed (see E in
) on a byte-by-byte basis in the direction specified by the R/W bit sent by the calling master.
Data can be changed only while I2C_SCL is low and must be held stable while I2C_SCL is high, as
shows. I2C_SCL is pulsed once for each data bit, with the msb being sent first. The receiving device must acknowledge each byte by pulling I2C_SDA low at the ninth clock; therefore, a data byte
transfer takes nine clock pulses. See Figure 27-3 .
I2C_SCL
I2C_SCL Held Low while
Interrupt is Serviced
1 2 3 4 5 6 7 8 9
I2C_SDA
START
Signal
Bit7 Bit6 Bit5 Bit4 Bit3 Bit2 Bit1
Bit0
R/W
Slave Address
ACK from
Receiver
1 2 3 4 5 6 7 8 9
Interrupt Bit Set
(Byte Complete)
Bit7 Bit6 Bit5 Bit4 Bit3 Bit2 Bit1 Bit0
Data Byte
No
ACK
Bit
STOP
Signal
Figure 27-3. Data Transfer
27.4.4
Acknowledge
The transmitter releases the I2C_SDA line high during the acknowledge clock pulse as shown in
. The receiver pulls down the I2C_SDA line during the acknowledge clock pulse so that it remains stable low during the high period of the clock pulse.
If it does not acknowledge the master, the slave receiver must leave I2C_SDA high. The master can then generate a STOP signal to abort the data transfer or generate a START signal (repeated start, shown in
Figure 27-5 and discussed in Section 27.4.6, “Repeated START”
) to start a new calling sequence.
27-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
I
2
C Interface
I2C_SCL
I2C_SDA by Transmitter
I2C_SDA by Receiver
1 2 3 4 5 6 7 8 9
Bit7 Bit6 Bit5 Bit4 Bit3 Bit2 Bit1
Bit0
R/W
START Signal
Figure 27-4. Acknowledgement by Receiver
ACK
If the master receiver does not acknowledge the slave transmitter after a byte transmission, it means end-of-data to the slave. The slave releases I2C_SDA for the master to generate a STOP or START signal
(
).
27.4.5
STOP Signal
The master can terminate communication by generating a STOP signal to free the bus. A STOP signal is defined as a low-to-high transition of I2C_SDA while I2C_SCL is at logical high (see F in
).
The master can generate a STOP even if the slave has generated an acknowledgment, at which point the slave must release the bus. The master may also generate a START signal following a calling address, without first generating a STOP signal. Refer to
Section 27.4.6, “Repeated START.”
27.4.6
Repeated START
A repeated START signal is a START signal generated without first generating a STOP signal to terminate the communication. The master uses a repeated START to communicate with another slave or with the same slave in a different mode (transmit/receive mode) without releasing the bus.
I2C_SCL msb
1 2 3 4 5 lsb
6 7 8 9 msb
1 2 3 4 5 lsb
6 7 8 9
I2C_SDA AD7 AD6 AD5 AD4 AD3 AD2 AD1 R/W XX AD7 AD6 AD5 AD4 AD3 AD2 AD1 R/W
START
Signal
Calling Address
R/W ACK
Bit
Repeated
START
Signal
A
New Calling Address
Stop
Figure 27-5. Repeated START
R/W No
ACK
Bit
STOP
Signal
Various combinations of read/write formats are then possible:
• The first example in
is the case of master-transmitter transmitting to slave-receiver.
The transfer direction is not changed.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
27-5
I
2
C Interface
•
The second example in Figure 27-6
is the master reading the slave immediately after the first byte.
At the moment of the first acknowledge, the master-transmitter becomes a master-receiver and the slave-receiver becomes slave-transmitter.
•
In the third example in Figure 27-6 , the START condition and slave address are both repeated using
the repeated START signal. This is to communicate with same slave in a different mode without releasing the bus. The master transmits data to the slave first, and then the master reads data from slave by reversing the R/W bit.
ST = Start
SP = Stop
A = Acknowledge (I2C_SDA low)
A = Not Acknowledge (I2C_SDA high)
Rept ST = Repeated Start
From Master to Slave
From Slave to Master
Example 1:
ST 7bit Slave Address
R/W
0 A Register Address A DATA A/A SP
Example 2:
ST 7bit Slave Address
Note: No acknowledge on the last byte
R/W
1 A DATA A DATA A SP
Example 3:
ST
7-bit Slave
Address
R/W
1 A DATA A
Rept
ST
R/W
7-bit Slave 0
Address
A DATA A DATA A/A SP
Master Reads from Slave Master Writes to Slave
Figure 27-6. Data Transfer, Combined Format
27.4.7
Clock Synchronization and Arbitration
I
2
C is a true multi-master bus that allows more than one master to be connected to it. If two or more master devices simultaneously request control of the bus, a clock synchronization procedure determines the bus clock. Because wire-AND logic is performed on the I2C_SCL line, a high-to-low transition on the
I2C_SCL line affects all the devices connected on the bus. The devices start counting their low period and once a device’s clock has gone low, it holds the I2C_SCL line low until the clock high state is reached.
However, the change of low to high in this device’s clock may not change the state of the I2C_SCL line if another device clock is still within its low period. Therefore, synchronized clock I2C_SCL is held low by the device with the longest low period.
Devices with shorter low periods enter a high wait state during this time (see
). When all devices concerned have counted off their low period, the synchronized clock (I2C_SCL) line is released and pulled high. At this point, the device clocks and the I2C_SCL line are synchronized, and the devices start counting their high periods. The first device to complete its high period pulls the I2C_SCL line low again.
27-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
I
2
C Interface
The relative priority of the contending masters is determined by a data arbitration procedure. A bus master loses arbitration if it transmits logic 1 while another master transmits logic 0. The losing masters
case the transition from master to slave mode does not generate a STOP condition. Meanwhile, hardware sets I2SR[IAL] to indicate loss of arbitration.
I2C_SCL
I2C_SDA by
Master1
I2C_SDA by
Master2
I2C_SDA
Master 2 Loses Arbitration, and becomes slave-receiver
Figure 27-7. Arbitration Procedure
Wait
Start counting high period
I2C_SCL1
I2C_SCL2
I2C_SCL
Internal Counter Reset
Figure 27-8. Clock Synchronization
27.4.8
Handshaking and Clock Stretching
The clock synchronization mechanism can be used as a handshake in data transfers. Slave devices can hold
I2C_SCL low after completing one byte transfer. In such a case, the clock mechanism halts the bus clock and forces the master clock into wait states until the slave releases I2C_SCL.
Slaves may also slow down the transfer bit rate. After the master has driven I2C_SCL low, the slave can drive I2C_SCL low for the required period and then release it. If the slave I2C_SCL low period is longer than the master I2C_SCL low period, the resulting I2C_SCL bus signal low period is stretched.
27.5
Memory Map/Register Definition
lists the configuration registers used in the I
2
C interface.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
27-7
I
2
C Interface
Table 27-1. I
2
C Module Memory Map
IPSBAR
Offset
Register
0x00_0300 I
2
C Address Register (I2ADR)
0x00_0304 I
2
C Frequency Divider Register (I2FDR)
0x00_0308 I
2
C Control Register (I2CR)
0x00_030C I
2
C Status Register (I2SR)
0x00_0310 I
2
C Data I/O Register (I2DR)
Access Reset Value Section/Page
R/W
R/W
R/W
R/W
R/W
0x00
0x00
0x00
0x81
0x00
27.5.1
I
2
C Address Register (I2ADR)
The I2ADR holds the address the I
2
C responds to when addressed as a slave. Note that it is not the address sent on the bus during the address transfer when the module is performing a master transfer.
Address: 0x00_0300 (I2ADR)
7 6
R
W
Reset: 0
5 4
ADR
3 2
0 0 0 0
Figure 27-9. I
2
C Address Register (I2ADR)
0
Table 27-2. I2ADR Field Descriptions
Access: User read/write
1 0
0
0 0
Field
7–1
ADR
0
Description
Slave address. Contains the specific slave address to be used by the I
2
C module. Slave mode is the default I
2
C mode for an address match on the bus.
Reserved, should be cleared.
27.5.2
I
2
C Frequency Divider Register (I2FDR)
The I2FDR, shown in Figure 27-10 , provides a programmable prescaler to configure the I
2
C clock for bit-rate selection.
Access: User read/write
1 0
Address: 0x00_0304 (I2FDR)
7
0 R
W
Reset: 0
6
0
5 4 3
IC
2
0 0 0 0 0
Figure 27-10. I
2
C Frequency Divider Register (I2FDR)
0 0
27-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
I
2
C Interface
Table 27-3. I2FDR Field Descriptions
Field
7–6
5–0
IC
Description
Reserved, should be cleared.
I
2
C clock rate. Prescales the clock for bit-rate selection. The serial bit clock frequency is equal to the internal bus clock divided by the divider shown below. Due to potentially slow I2C_SCL and I2C_SDA rise and fall times, bus signals are sampled at the prescaler frequency.
IC
0x08
0x09
0x0A
0x0B
0x0C
0x0D
0x0E
0x0F
0x00
0x01
0x02
0x03
0x04
0x05
0x06
0x07
Divider
144
160
192
240
80
88
104
128
44
48
56
68
28
30
34
40
IC
0x18
0x19
0x1A
0x1B
0x1C
0x1D
0x1E
0x1F
0x10
0x11
0x12
0x13
0x14
0x15
0x16
0x17
Divider IC
1152
1280
1536
1920
2304
2560
3072
3840
576
640
768
960
288
320
384
480
0x28
0x29
0x2A
0x2B
0x2C
0x2D
0x2E
0x2F
0x20
0x21
0x22
0x23
0x24
0x25
0x26
0x27
Divider
80
96
112
128
48
56
64
72
28
32
36
40
20
22
24
26
IC
0x38
0x39
0x3A
0x3B
0x3C
0x3D
0x3E
0x3F
0x30
0x31
0x32
0x33
0x34
0x35
0x36
0x37
Divider
640
768
896
1024
1280
1536
1792
2048
320
384
448
512
160
192
224
256
27.5.3
I
2
C Control Register (I2CR)
The I2CR is used to enable the I
2
C module and the I
2
C interrupt. It also contains bits that govern operation as a slave or a master.
Address: 0x00_0308 (I2CR)
7
R
W
Reset:
IEN
0
6
IIEN
5
MSTA
4
MTX
3
TXAK
2
RSTA
0 0 0 0
Figure 27-11. I
2
C Control Register (I2CR)
0
Access: User read/write
1
0
0
0
0 0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
27-9
I
2
C Interface
Table 27-4. I2CR Field Descriptions
Field
7
IEN
6
IIEN
5
MSTA
4
MTX
3
TXAK
2
RSTA
1–0
Description
I
2
C enable. Controls the software reset of the entire I
2
C module. If the module is enabled in the middle of a byte transfer, slave mode ignores the current bus transfer and starts operating when the next START condition is detected.
Master mode is not aware that the bus is busy; so initiating a start cycle may corrupt the current bus cycle, ultimately causing either the current master or the I
2
C module to lose arbitration, after which bus operation returns to normal.
0 The I
2
C module is disabled, but registers can still be accessed.
1 The I
2
C module is enabled. This bit must be set before any other I2CR bits have any effect.
I
2
C interrupt enable.
0 I
2
C module interrupts are disabled, but currently pending interrupt condition is not cleared.
1 I
2
C module interrupts are enabled. An I
2
C interrupt occurs if I2SR[IIF] is also set.
Master/slave mode select bit. If the master loses arbitration, MSTA is cleared without generating a STOP signal.
0 Slave mode. Changing MSTA from 1 to 0 generates a STOP and selects slave mode.
1 Master mode. Changing MSTA from 0 to 1 signals a START on the bus and selects master mode.
Transmit/receive mode select bit. Selects the direction of master and slave transfers.
0 Receive
1 Transmit. When the device is addressed as a slave, software should set MTX according to I2SR[SRW]. In master mode, MTX should be set according to the type of transfer required. Therefore, when the MCU addresses a slave device, MTX is always 1.
Transmit acknowledge enable. Specifies the value driven onto I2C_SDA during acknowledge cycles for both master and slave receivers. Note that writing TXAK applies only when the I
2
C bus is a receiver.
0 An acknowledge signal is sent to the bus at the ninth clock bit after receiving one byte of data.
1 No acknowledge signal response is sent (that is, acknowledge bit = 1).
Repeat start. Always read as 0. Attempting a repeat start without bus mastership causes loss of arbitration.
0 No repeat start
1 Generates a repeated START condition.
Reserved, should be cleared.
27.5.4
I
2
C Status Register (I2SR)
This I2SR contains bits that indicate transaction direction and status.
Address: 0x00_030C (I2SR)
7
ICF R
W
Reset: 1
6
IAAS
5
IBB
4
IAL
3
0
0 0 0 0
Figure 27-12. I
2
C Status Register (I2SR)
2
SRW
0
Access: User read/write
1 0
RXAK
IIF
0 1
27-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
I
2
C Interface
Table 27-5. I2SR Field Descriptions
Field
7
ICF
6
IAAS
5
IBB
4
IAL
3
2
SRW
1
IIF
0
RXAK
Description
I
2
C Data transferring bit. While one byte of data is transferred, ICF is cleared.
0 Transfer in progress
1 Transfer complete. Set by the falling edge of the ninth clock of a byte transfer.
I
2
C addressed as a slave bit. The CPU is interrupted if I2CR[IIEN] is set. Next, the CPU must check SRW and set its TX/RX mode accordingly. Writing to I2CR clears this bit.
0 Not addressed.
1 Addressed as a slave. Set when its own address (IADR) matches the calling address.
I
2
C bus busy bit. Indicates the status of the bus.
0 Bus is idle. If a STOP signal is detected, IBB is cleared.
1 Bus is busy. When START is detected, IBB is set.
I
2
C arbitration lost. Set by hardware in the following circumstances. (IAL must be cleared by software by writing zero to it.)
• I2C_SDA sampled low when the master drives high during an address or data-transmit cycle.
• I2C_SDA sampled low when the master drives high during the acknowledge bit of a data-receive cycle.
• A start cycle is attempted when the bus is busy.
• A repeated start cycle is requested in slave mode.
• A stop condition is detected when the master did not request it.
Reserved, should be cleared.
Slave read/write. When IAAS is set, SRW indicates the value of the R/W command bit of the calling address sent from the master. SRW is valid only when a complete transfer has occurred, no other transfers have been initiated, and the I
2
C module is a slave and has an address match.
0 Slave receive, master writing to slave.
1 Slave transmit, master reading from slave.
I
2
C interrupt. Must be cleared by software by writing a zero in the interrupt routine.
0 No I
2
C interrupt pending
1 An interrupt is pending, which causes a processor interrupt request (if IIEN = 1). Set when one of the following occurs:
• Complete one byte transfer (set at the falling edge of the ninth clock)
• Reception of a calling address that matches its own specific address in slave-receive mode
• Arbitration lost
Received acknowledge. The value of I2C_SDA during the acknowledge bit of a bus cycle.
0 An acknowledge signal was received after the completion of 8-bit data transmission on the bus
1 No acknowledge signal was detected at the ninth clock.
27.5.5
I
2
C Data I/O Register (I2DR)
In master-receive mode, reading the I2DR allows a read to occur and for the next data byte to be received.
In slave mode, the same function is available once the I
2
C has received its slave address.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
27-11
I
2
C Interface
Address: 0x00_0310 (I2DR)
7
R
W
Reset: 0
6 5 4
DATA
3 2
0 0 0 0
Figure 27-13. I
2
C Data I/O Register (I2DR)
0
Access: User read/write
1 0
0 0
Field
7–0
DATA
Table 27-6. I2DR Field Description
Description
I
2
C data. In master transmit mode, when data is written to this register, a data transfer is initiated. The most significant bit is sent first. In master receive mode, reading this register initiates the reception of the next byte of data.
In slave mode, the same functions are available after an address match has occurred.
Note: In master transmit mode, the first byte of data written to I2DR following assertion of I2CR[MSTA] is used for the address transfer and should comprise the calling address (in position D7–D1) concatenated with the required R/W bit (in position D0). This bit (D0) is not automatically appended by the hardware, software must provide the appropriate R/W bit.
Note: I2CR[MSTA] generates a start when a master does not already own the bus. I2CR[RSTA] generates a start
(restart) without the master first issuing a stop (i.e., the master already owns the bus). In order to start the read of data, a dummy read to this register starts the read process from the slave. The next read of the I2DR register contains the actual data.
27.6
I
2
C Programming Examples
The following examples show programming for initialization, signaling START, post-transfer software response, signaling STOP, and generating a repeated START.
27.6.1
Initialization Sequence
Before the interface can transfer serial data, registers must be initialized:
1. Set I2FDR[IC] to obtain I2C_SCL frequency from the system bus clock. See Section 27.5.2, “I2C
Frequency Divider Register (I2FDR).”
2. Update the I2ADR to define its slave address.
3. Set I2CR[IEN] to enable the I
2
C bus interface system.
4. Modify the I2CR to select or deselect master/slave mode, transmit/receive mode, and interrupt-enable or not.
NOTE
If I2SR[IBB] is set when the I
2
C bus module is enabled, execute the following pseudocode sequence before proceeding with normal initialization code. This issues a STOP command to the slave device, placing it in idle state as if it were just power-cycled on.
I2CR = 0x0
I2CR = 0xA0 dummy read of I2DR
27-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
I
2
C Interface
I2SR = 0x0
I2CR = 0x0
27.6.2
Generation of START
After completion of the initialization procedure, serial data can be transmitted by selecting the master transmitter mode. On a multiple-master bus system, I2SR[IBB] must be tested to determine whether the serial bus is free. If the bus is free (IBB = 0), the START signal and the first byte (the slave address) can be sent. The data written to the data register comprises the address of the desired slave and the lsb indicates the transfer direction.
The free time between a STOP and the next START condition is built into the hardware that generates the
START cycle. Depending on the relative frequencies of the system clock and the I2C_SCL period, it may be necessary to wait until the I2C is busy after writing the calling address to the I2DR before proceeding with the following instructions.
The following example signals START and transmits the first byte of data (slave address):
CHFLAG MOVE.B I2SR,-(A0)
BTST.B #5, (A0)+
BNE.S CHFLAG
TXSTART MOVE.B I2CR,-(A0)
BSET.B #4,(A0)
MOVE.B (A0)+, I2CR
MOVE.B I2CR, -(A0)
BSET.B #5, (A0)
MOVE.B (A0)+, I2CR
IFREE
MOVE.B CALLING,-(A0)
MOVE.B (A0)+, I2DR
MOVE.B I2SR,-(A0)
;Check I2SR[MBB]
;If I2SR[MBB] = 1, wait until it is clear
;Set transmit mode
;Set master mode
;Generate START condition
;Transmit the calling address, D0=R/W
;Check I2SR[MBB]
;If it is clear, wait until it is set.
BTST.B #5, (A0)+
BEQ.S IFREE;
27.6.3
Post-Transfer Software Response
Sending or receiving a byte sets the I2SR[ICF], which indicates one byte communication is finished.
I2SR[IIF] is also set. An interrupt is generated if the interrupt function is enabled during initialization by setting I2CR[IIEN]. Software must first clear I2SR[IIF] in the interrupt routine. I2SR[ICF] is cleared either by reading from I2DR in receive mode or by writing to I2DR in transmit mode.
Software can service the I2C I/O in the main program by monitoring the IIF bit if the interrupt function is disabled. Polling should monitor IIF rather than ICF, because that operation is different when arbitration is lost.
When an interrupt occurs at the end of the address cycle, the master is always in transmit mode; that is, the address is sent. If master receive mode is required, I2CR[MTX] should be toggled.
During slave-mode address cycles (I2SR[IAAS] = 1), I2SR[SRW] is read to determine the direction of the next transfer. MTX is programmed accordingly. For slave-mode data cycles (IAAS = 0), SRW is invalid.
MTX should be read to determine the current transfer direction.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
27-13
I
2
C Interface
The following is an example of a software response by a master transmitter in the interrupt routine (see
).
I2SR LEA.L I2SR,-(A7)
BCLR.B #1,(A7)+
MOVE.B I2CR,-(A7)
BTST.B #5,(A7)+
BEQ.S SLAVE
MOVE.B I2CR,-(A7)
BTST.B #4,(A7)+
BEQ.S RECEIVE
MOVE.B I2SR,-(A7)
BTST.B #0,(A7)+
BNE.B END
TRANSMIT MOVE.B DATABUF,-(A7)
MOVE.B (A7)+, I2DR
;Load effective address
;Clear the IIF flag
;Push the address on stack,
;check the MSTA flag
;Branch if slave mode
;Push the address on stack
;check the mode flag
;Branch if in receive mode
;Push the address on stack,
;check ACK from receiver
;If no ACK, end of transmission
;Stack data byte
;Transmit next byte of data
27.6.4
Generation of STOP
A data transfer ends when the master signals a STOP, which can occur after all data is sent, as in the following example.
MASTX ;If no ACK, branch to end
END
MOVE.B I2SR, -(A7)
BTST.B #0,(A7)+
BNE.B END
MOVE.B TXCNT,D0
BEQ.S END
MOVE.B DATABUF,-(A7)
MOVE.B (A7)+,I2DR
MOVE.B TXCNT,D0
SUBQ.L #1,D0
MOVE.B D0,TXCNT
BRA.S EMASTX;Exit
LEA.L I2CR,-(A7)
BCLR.B #5,(A7)+
EMASTX RTE
;Get value from the transmitting counter
;If no more data, branch to end
;Transmit next byte of data
;Decrease the TXCNT
;Generate a STOP condition
;Return from interrupt
For a master receiver to terminate a data transfer, it must inform the slave transmitter by not acknowledging the last data byte. This is done by setting I2CR[TXAK] before reading the next-to-last byte. Before the last byte is read, a STOP signal must be generated, as in the following example.
MASR MOVE.B RXCNT,D0
SUBQ.L #1,D0
MOVE.B D0,RXCNT
BEQ.S ENMASR
MOVE.B RXCNT,D1
EXTB.L D1
SUBI.L #1,D1;
BNE.S NXMAR
LAMAR BSET.B #3,I2CR
BRA NXMAR
ENMASR BCLR.B #5,I2CR
NXMAR MOVE.B I2DR,RXBUF
;Decrease RXCNT
;Last byte to be read
;Check second-to-last byte to be read
;Not last one or second last
;Disable ACK
;Last one, generate STOP signal
;Read data and store RTE
27-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
I
2
C Interface
27.6.5
Generation of Repeated START
After the data transfer, if the master still wants the bus, it can signal another START followed by another slave address without signaling a STOP, as in the following example.
;Repeat START (RESTART) RESTART MOVE.B I2CR,-(A7)
BSET.B #2, (A7)
MOVE.B (A7)+, I2CR
MOVE.B CALLING,-(A7)
MOVE.B CALLING,-(A7)
MOVE.B (A7)+, I2DR
;Transmit the calling address, D0=R/W-
27.6.6
Slave Mode
In the slave interrupt service routine, software should poll the I2SR[IAAS] bit to determine if the controller has received its slave address. If IAAS is set, software should set the transmit/receive mode select bit
(I2CR[MTX]) according to the I2SR[SRW]. Writing to the I2CR clears the IAAS automatically. The only time IAAS is read as set is from the interrupt at the end of the address cycle where an address match occurred; interrupts resulting from subsequent data transfers will have IAAS cleared. A data transfer can now be initiated by writing information to I2DR for slave transmits, or read from I2DR in slave-receive mode. A dummy read of I2DR in slave/receive mode releases I2C_SCL, allowing the master to send data.
In the slave transmitter routine, I2SR[RXAK] must be tested before sending the next byte of data. Setting
RXAK means an end-of-data signal from the master receiver, after which software must switch it from transmitter to receiver mode. Reading I2DR then releases I2C_SCL so that the master can generate a STOP signal.
27.6.7
Arbitration Lost
If several devices try to engage the bus at the same time, one becomes master. Hardware immediately switches devices that lose arbitration to slave receive mode. Data output to I2C_SDA stops, but I2C_SCL is still generated until the end of the byte during which arbitration is lost. An interrupt occurs at the falling edge of the ninth clock of this transfer with I2SR[IAL] = 1 and I2CR[MSTA] = 0.
If a device that is not a master tries to transmit or execute a START, hardware will inhibit the transmission, clear MSTA without signaling a STOP, generate an interrupt to the CPU, and set IAL to indicate a failed attempt to engage the bus. When considering these cases, the slave service routine should first test IAL and software should clear it if it is set.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
27-15
I
2
C Interface
Clear
IIF
Y Master
Mode?
N
TX TX/Rx
?
RX Y
Arbitration
Lost?
N
Last Byte
Transmitted
?
N
Y
Y
RXAK= 0
?
Y
N
End of
ADDR Cycle
(Master RX)
?
N
Write Next
Byte to I2DR
Last
Byte to be
Read ?
N
Y
Y
2nd Last
Byte to be
Read?
N
Set TXAK =1
Generate
STOP Signal
Clear IAL
N
IAAS=1
?
Address
Cycle
Y
(Read) Y
Y
SRW=1
?
N (WRITE)
IAAS=1
?
N
Data
Cycle
Tx/Rx
?
TX
RX
Set TX
Mode
Write Data to I2DR
Y
Tx Next
Byte
ACK from
Receiver
?
N
Read Data from I2DR and Store
Switch to
Rx Mode
Set RX
Mode
Switch to
Rx Mode
Dummy Read from I2DR
Generate
STOP Signal
Read Data from I2DR
And Store
Dummy Read from I2DR
Dummy Read from I2DR
RTE
Figure 27-14. Flow-Chart of Typical I
2
C Interrupt Routine
27-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 28
Analog-to-Digital Converter (ADC)
28.1
Introduction
The analog-to-digital converter (ADC) consists of two separate and complete ADCs, each with their own sample and hold circuits. The converters share a common voltage reference and common digital control module.
28.2
Features
The ADC’s characteristics include the following:
• 12-bit resolution
• Maximum ADC clock frequency of 5.0 MHz, 200ns period
• Sampling rate up to 1.66 million samples per second
1
• Single conversion time of 8.5 ADC clock cycles (8.5 × 200ns = 1.7
µ s)
• Additional conversion time of 6 ADC clock cycles (6 × 200ns = 1.2
µ s)
• Eight conversions in 26.5 ADC clocks (26.5
× 200ns = 5.3
µ s) using simultaneous mode
• Ability to simultaneously sample and hold 2 inputs
• Ability to sequentially scan and store up to 8 measurements
• Internal multiplex to select two of 8 inputs
• Power savings modes allow automatic shutdown/startup of all or part of ADC
• Those inputs not selected tolerate injected/sourced current without affecting ADC performance, supporting operation in noisy industrial environments.
• Optional interrupts at the end of a scan, if an out-of-range limit is exceeded (high or low), or at zero crossing
• Optional sample correction by subtracting a pre-programmed offset value
• Signed or unsigned result
• Single ended or differential inputs for all input pins with support for an arbitrary mix of input types
28.3
Block Diagram
The ADC function, shown in
consists of two four-channel input select functions, interfacing with two independent Sample and Hold (S/H) circuits, which feed two 12-bit ADCs. The two converters store their results in a buffer, awaiting further processing.
1. Once in loop mode, the time between each conversion is 6 ADC clock cycles (1.2
µ s at 5.0 MHz). Using simultaneous conversion, two samples are captured in 1.2
µ s, providing an overall sample rate of 1.66 million samples per second.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-1
Analog-to-Digital Converter (ADC)
V
REFH
V
REFL
Voltage
Reference
Circuit
Scaling & Cyclic
Converter A
12
Scaling & Cyclic
Converter B
12
Digital Output
Storage Registers
AN0
AN1
AN2
AN3
AN4
AN5
AN6
AN7
MUX
MUX
Sample/Hold
16
SYNCx Controller Bus Interface
IRQ
Data
Figure 28-1. Dual ADC Block Diagram
28.4
Functional Description
The ADC’s conversion process is either initiated by a sync signal from one of two input pins (SYNCx) or by writing 1 to a START n bit.
Starting a single conversion actually begins a sequence of conversions, or a scan of up to 8 single-ended or differential samples one at a time in sequential scan mode. The operation of the module in sequential
scan mode is shown in Figure 28-2 .
28-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
HILIM[0:3]
>
ADC0
ADC1
ADC2
LOLIM[0:3]
AN0
AN1
AN2
AN3
V
REFL
V+
V–
ADCA
Test Data
(From CPU)
12
12
Zero Crossing Logic
<
12
+
–
+
13
OFFST[0:3]
RSLT[0:3]
End of
Scan A
Interrupt
Channel Select
Single-Ended vs
Differential
Zero Crossing or Error Limit
Interrupt
HILIM[4:7]
AN4
AN5
AN6
AN7
V
REFL
Crossbars allow AN0-3 to be stored in samples 4-7, or AN4-7 to be stored in samples 0-3
LOLIM[4:7]
>
<
Zero Crossing Logic
V+
V–
ADCB
12
12
Channel Select
Single-Ended vs
Differential
Test Data
(From CPU)
12
+
–
+
13
OFFST[4:7]
RSLT[4:7]
Figure 28-2. Sequential Mode Operation of the ADC
End of
Scan B
Interrupt
Scan sequence is determined by defining 8 sample slots in ADLST1/2 registers, processed in order
SAMPLE0-7 during sequential scan, or in order SAMPLE0-3 by converter A and in order SAMPLE4-7 by converter B in parallel scan. SAMPLE slots may be disabled using the SDIS register.
The following pairs of analog inputs can be configured as a differential pair: AN0-1, AN2-3, AN4-5, and
AN6-7. When configured as a differential pair, a reference to either member of the differential pair by a sample slot results in a differential measurement using that differential pair.
Parallel scan can be simultaneous or non-simultaneous. During simultaneous scan, the scans in the two converters are done simultaneously and always result in simultaneous pairs of conversions, one by converter A and one by converter B. The two converters share the same start, stop, sync, end-of-scan interrupt enable control, and interrupt. Scanning in both converters is terminated when either converter encounters a disabled sample. In non-simultaneous scan, the parallel scans in the two converters are achieved independently. The two converters have their own start, stop, sync, end-of-scan interrupt enable controls, and end-of-scan interrupts. Scanning in either converter terminates only when that converter encounters a disabled sample in its part of SDIS register (DS0-DS3 for A, DS4-DS7 for B).
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-3
Analog-to-Digital Converter (ADC)
HILIM[0:3]
>
ADC0
ADC1
ADC2
LOLIM[0:3]
AN0
AN1
AN2
AN3
V
REFL
V+
V–
ADCA
Test Data
(From CPU)
12
12
Zero Crossing Logic
<
12
+
–
+
13
OFFST[0:3]
RSLT[0:3]
End of
Scan A
Interrupt
Channel Select
Single-Ended vs
Differential
Zero Crossing or Error Limit
Interrupt
HILIM[4:7]
Crossbars do not operate in this mode
LOLIM[4:7]
>
AN4
AN5
AN6
AN7
V
REFL
End of
Scan B
Interrupt
<
Zero Crossing Logic
V+
V–
ADCB
12
12
Channel Select
Single-Ended vs
Differential
Test Data
(From CPU)
12
+
–
+
13
OFFST[4:7]
RSLT[4:7]
Figure 28-3. Parallel Mode Operation of the ADC
The ADC can be configured to perform a single scan and halt, perform a scan whenever triggered, or perform the scan sequence repeatedly until manually stopped. The single scan (once mode) differs from the triggered mode only in that SYNC input signals must be re-armed after each using a once mode scan, and subsequent SYNC inputs are ignored until the SYNC input is re-armed. This arming can occur anytime after the SYNC pulse occurs, even while the scan it initiated is still in process.
Optional interrupts can be generated at the end of a scan sequence. Interrupts are available simply to indicate the scan ended, that a sample was out of range, or at several different zero crossing conditions.
Out-of-range is determined by the high and low limit registers.
To understand the operation of the ADC, it is important to understand the features and limitations of each of the functional parts.
28-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
28.4.1
Input MUX Function
The input MUX function is shown in Figure 28-4 .
The channel select and single ended vs. differential switches are indirectly controlled based on settings within the LIST1, LIST2, and SDIS registers, and the
CHNCFG field of the CTRL1 register.
1. MUXing for Sequential mode, single-ended conversions—During each conversion cycle (sample), any one input of the two muxes can be directed to any RSLT n register.
2. MUXing for sequential mode, differential conversions—During any conversion cycle (sample), either member of a differential pair may be referenced as a SAMPLE, resulting in a differential measurement on that pair being stored in the corresponding RSLT n register.
3. MUXing for parallel mode, single-ended conversions—During any conversion cycle (sample), any of AN0-AN3 can be directed to an RSLT0-3 Result register and any of AN4-AN7 can be directed to the RSLT4-7.
4. MUXing for parallel mode, differential conversions—During any conversion cycle (sample), either member of differential pair AN0/1 or either member of differential pair AN2/3 can be referenced as a SAMPLE, resulting in a differential measurement of that pair being stored in one of the RSLT0-3 registers. Likewise, either member of differential pair AN4/5 or either member of differential pair AN6/7 can be referenced as a SAMPLE, resulting in a differential measurement of that pair being stored in one of the RSLT4-7 registers.
Details of switch operation is shown in Table 28-2 .
Internally, all measurements are performed differentially. During single ended measurements, V
REFL the selected analog input is used as the positive ( + ) input.
is used as the negative (-) input voltage, while
Table 28-2. Analog MUX Controls for Each Conversion Mode
Conversion Mode
Sequential, Single Ended
Sequential, Differential
Parallel, Single Ended
Parallel, Differential
Channel Select Switches Single Ended Differential Switches
The two 1-of-4 select muxes can be set for the appropriate input line.
The lower switch selects V
REFL
for the
V- input of the A/D. The upper switch is always closed so that any of the four inputs can get to the V+ A/D input.
The channel select switches are turned on in pairs, providing a dual
1-of-2 select function, such that either of the two differential channels can be routed to the A/D input.
The upper switch is open and the bottom switch selects the differential channel for the V- input of the A/D.
The two 1-of-4 select muxes can be set for the appropriate input line.
The lower switch selects V
REFL
for the
V- input of the A/D. The upper switch is always closed so that any of the four inputs can get to the V+ A/D input.
The channel select switches are turned on in pairs, providing a dual
1-of-2 select function, such that either of the two differential channels can be routed to the A/D input.
The upper and lower switches are open and the middle switch is closed, providing the differential channel to the differential input of the A/D.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-5
Analog-to-Digital Converter (ADC)
MUX Configuration for Single-Ended vs MUX Configuration for Differential
AN0
AN1
AN2
AN3
V
REFL
Channel
Select
V+
V–
Single-Ended
Channel Select
Single-Ended vs
Differential
To
Converter A
Interface
Function
AN0
AN1
AN2
AN3
V
REFL
Channel
Select
V+
V–
Differential
Channel Select
Single-Ended vs
Differential
To
Converter A
Interface
Function
AN4
AN5
AN6
AN7
V
REFL
Channel
Select
Channel
Select
V+
V–
To
Converter B
Interface
Function
AN4
AN5
AN6
AN7
V
REFL
Single-Ended
Channel Select
Single-Ended vs
Differential
Figure 28-4. Input Select Mux
V+
V–
Differential
Channel Select
Single-Ended vs
Differential
To
Converter B
Interface
Function
28.4.2
ADC Sample Conversion
The ADC consists of a cyclic, algorithmic architecture using two recursive sub-ranging sections (RSD#1
and RSD#2), shown in Figure 28-5 .
Each sub-ranging section resolves a single bit for each conversion clock, resulting in an overall conversion rate of two bits per clock cycle. Each sub-ranging section is designed to run at a maximum clock speed of 5.0 MHz. Thus a complete 12-bit conversion takes 6 ADC clocks (1.2ms), not including sample or post processing time.
28-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
ADCA
Cyclic ADC Core
AN0
AN1
AN2
AN3
V
REFL
V+
V–
V+
V–
Channel Select
Single-Ended vs
Differential
Interface
Function
MUX
RSD#1
θ
1
RSD#2
θ
2
ADCB
Cyclic ADC Core
AN4
AN5
AN6
AN7
V
REFL
V+
V–
V+
V–
Interface
Function
MUX
RSD#1
θ
1
RSD#2
θ
2
Channel Select
Single-Ended vs
Differential
Figure 28-5. Cyclic ADC — Top Level Block Diagram
The input mode for a given sample is determined by the CHNCFG field of the CTRL1 register. The ADC has two input modes:
1. Single-ended mode (CHNCFG bit=0)—In single-ended mode, input mux of the ADC selects one of the analog inputs and directs it to the plus terminal of the A/D core. The minus terminal of the
A/D core is connected to the V
REFL
reference during this mode. The ADC measures the voltage of the selected analog input and compares it against the (V
REFH
- V
REFL
) reference voltage range.
2. Differential mode (CHNCFG bit = 1)—In differential mode, the ADC measures the voltage difference between two analog inputs and compares that against the (V
REFH
- V
REFL
) voltage range. The input is selected as an input pair: AN0/1, AN2/3, AN4/5, or AN6/7. In this mode, the plus terminal of the A/D core is connected to the even analog input, while the minus terminal is connected to the odd analog input.
A mix and match combination of differential and single-ended configurations may exist.
Examples:
• AN0 and AN1 differential, AN2 and AN3 single-ended
• AN4 and AN5 differential, AN6 and AN7 single-ended
28.4.2.1
Single-Ended Samples
The ADC module performs a ratio metric conversion. For single ended measurements, the digital result is proportional to the ratio of the analog input to the reference voltage in the following formula:
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-7
Analog-to-Digital Converter (ADC)
SingleEndedValue
= round (
V
V IN –
REFH
V
–
REFLO
V REFLO
× 4095 )
× 8
VIN = Applied voltage at the input pin
VREFH and VREFL = Voltage at the external reference pins on the device (typically VREFH = VSSA and
VREFL = VDDA)
Note: The 12-bit result is rounded to the nearest LSB.
Note: The ADC is a 12-bit function with 4096 possible states. However, the 12 bits have been left shifted three bits on the 16-bit data bus so its magnitude, as read from the data bus, is now
32760.
28.4.2.2
Differential Samples
For differential measurements, the digital result is proportional to the ratio of the difference in the inputs to the difference in the reference voltages (V
REFH for differential inputs.
and V
REFL
shows typical configurations
When converting differential measurements, the following formula is useful:
DifferentialValue
= round (
V
V IN 1 V
REFH – V
IN 2
REFLO
× 4095 )
× 8
V
IN
= Applied voltage at the input pin
V
REFH and V
REFL
= Voltage at the external reference pins on the device (typically V
REFH
= V
SSA
and V
REFL
=
V
DDA
)
Note: The 12-bit result is rounded to the nearest LSB.
Note: The ADC is a 12-bit function with 4096 possible states. However, the 12 bits have been left shifted three bits on the 16-bit data bus so its magnitude, as read from the data bus, is now
32760.
28-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
V
REFH
Potential
AN+
+
–
AN–
Differential buffer will center about mid-point
AN+
NOTE: Normally, V
REFL set to V
SSA
= 0V
is
AN–
V
REF
/2
Center tap held at (V
REFH
+ V
REFL
) /2
Figure 28-6. Typical Connections for Differential Measurements
28.4.3
ADC Data Processing
correction. The adder subtracts the ADOFS n register value from each sample and the result is stored in the corresponding result register (RSLT n ). Concurrent to this the raw ADC value is checked for limit violations, and the RSLT n values are checked for zero-crossing. Appropriate interrupts are asserted, if enabled.
The sign of the result is calculated from the ADC unsigned result minus the respective offset register. If the offset register is programmed with a value of zero, the result register value is unsigned and equals the cyclic converter unsigned result. The range of the result (RSLT) register is $0000–$7FF8, assuming the offset (ADOFS n ) register is set to zero.
The processor can write to the result registers whenever the ADC is in stop mode or powered down. The data from this write operation is treated as if it came from the ADC analog core; so the limit checking, zero crossing, and the offset registers function as if in normal mode. For example, if the ADC is stopped and the processor writes to RSLT5 register, the data written to the RSLT5 register is muxed to the ADC digital logic inputs, processed, and stored into RSLT5, as if the analog core had provided the data. This test data must be left justified by 3 bits (as shown in the RSLT register definition) and does not include the sign bit.
The sign bit (SEXT) is calculated during subtraction of the corresponding ADOFS n offset value.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-9
Analog-to-Digital Converter (ADC)
V+
V–
ADCA
Test Data
(From CPU)
12
12
HILIM[0:3]
LOLIM[0:3]
Zero Crossing Logic
<
12
+
–
+
13
OFFST[0:3]
>
RSLT[0:3]
End of
Scan A
Interrupt
ADC0
ADC1
ADC2
Zero Crossing or Error Limit
Interrupt
V+
V–
ADCB
12
12
HILIM[4:7]
>
LOLIM[4:7]
Zero Crossing Logic
<
12
+
–
+
13
OFFST[4:7]
RSLT[4:7]
Test Data
(From CPU)
Figure 28-7. Result Register Data Manipulation
End of
Scan B
Interrupt
28.4.4
Sequential vs. Parallel Sampling
All scan modes make use of the 8 SAMPLE slots in the ADLST1 and ADLST2 registers. These slots are used to define which single-ended input or differential input pair is measured at each step in a scan sequence. The SDIS register is used to disable unneeded slots.
Differential measurements are made on input pairs AN0/1, AN2/3, AN4/5, and AN6/7 using the CHNCFG field of the CTRL1 register. A single ended measurement will be made if a SAMPLE slot refers to an input not configured as a member of a differential pair by CHNCFG. A differential measurement will be made if a SAMPLE slot refers to either member of a differential pair. Refer to the CHNCFG field description in the CTRL1 register for details of differential and single ended measurement.
Scan modes are either sequential or parallel, as defined by the SMODE field of the CTRL1 register. In sequential scans, up to 8 SAMPLE slots are sampled one at a time in the order SAMPLE 0-7. Each
SAMPLE slot may refer to any of the 8 analog inputs (AN0-7), thus the same input may be referenced by
28-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC) more than one SAMPLE slot. Scanning is initiated when the START0 bit is written as 1 or, if the SYNC0 bit is 1, when the SYNC0 input goes high. A scan ends when the first disabled sample slot is encountered in the SDIS register. Completion of the scan triggers the EOSI0 interrupt if the interrupt is enabled by the
EOSIE0 bit. The START0 bit and SYNC0 input are ignored while a scan is in process. Scanning stops and cannot be initiated when the STOP0 bit is set.
Parallel scans differ in that converter A collects up to 4 samples (SAMPLE 0-3) in parallel to converter B collecting up to 4 samples (SAMPLE 4-7). SAMPLEs 0-3 may only reference inputs AN0-3, and
SAMPLEs 4-7 may only reference inputs AN4-7. Within these constraints, any sample may reference any pin and the same input may be referenced by more than one sample slot.
By default (when SIMULT=1), parallel scans of the converters are initiated together when the START0 bit is written as 1 or, if the SYNC0 bit is 1, when the SYNC0 input goes high. The scan in both converters terminates when either converter encounters a disabled sample slot in SDIS. Completion of a scan triggers the EOSI0 interrupt provided the EOSIE0 interrupt enable is set. Samples are always taken simultaneously in both the A and B converters. Setting the STOP0 bit stops and prevents the initiation of scanning in both converters.
Setting SIMULT=0 (non-simultaneous mode) causes parallel scanning to operate independently in the A and B converter. Each converter has its own set of START n , STOP n , SYNC n , and EOSIE n control bits,
SYNC n input, EOSI n interrupt, and CIP n status indicators ( n = 0 for converter A, n = 1 for converter B).
Though still operating in parallel, the scans in the A and B converter start and stop independently according to their own controls. They may be simultaneous, phase shifted, or asynchronous, depending on when scans are initiated on the respective converters. The A and B converter may be of different length
(still up to a maximum of four) and each converter’s scan completes when a disabled sample is encountered in that converters sample list only. STOP0 only stops the A converter, and STOP1 only stops the B converter. Looping scan modes repeat independently, with the A converter capturing SAMPLE 0-3, and B converter capturing SAMPLE 4-7. In loop modes, each converter independently restarts its scan after capturing its samples.
28.4.5
Scan Sequencing
Scan modes break down into three types based on how they repeat: once, triggered, or loop. Be certain to
read Section 28.4.4 “Sequential vs. Parallel Sampling
” to understand the operation of sequential and parallel scan modes before proceeding.
During a once mode scan, a single sequential or parallel scan is executed. Once scan modes differ from triggered scan modes in that they must be re-armed after each use. While all scan modes ignore sync pulses occurring while a scan is in process, once scan modes will continue to ignore sync pulses even after the scan completes until re-armed. However, re-arming can occur any time, including during the scan, by writing to a CTRL n register. If operating in a sequential mode or simultaneous parallel, write to the CTRL1 register. If operating in a non-simultaneous parallel mode, re-arm converter A by writing to the CTRL1 register and converter B by writing to the CTRL2 register.
Triggered scan modes are identical to the corresponding once scan modes, except that re-arming of sync inputs is not necessary.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-11
Analog-to-Digital Converter (ADC)
Loop scan modes automatically restart a scan as soon as the previous scan completes. In the loop sequential mode, up to 8 samples are captured in each loop, and the next scan starts immediately after the completion of the previous scan. In loop parallel scan modes, both converters restart together if
SIMULT=1 and restart independently if SIMULT=0. All subsequent start and sync pulses will be ignored after the scan begins. Scanning can only be terminated by setting a STOP n bit. Use STOP0 in the CTRL1 register if operating in a sequential or simultaneous parallel mode. If operating in a non-simultaneous parallel mode, use STOP0 to stop converter A and STOP1 in the CTRL2 register to stop converter B.
28.4.6
Power Management
The five supported power modes are described below. They are in order of highest to lowest power
” for details of the various clocks referenced below.
28.4.6.1
Power Management Modes
1. Normal power mode
This mode operates when:
— At least one ADC converter is powered up (PD0 or PD1=0 in the POWER register);
— Both auto power-down and auto standby modes are disabled (APD=0, ASB=0 in
ADCPOWER);
— The ADC’s clock is enabled (ADC=1 in the SIM module’s SIM_PCE register).
In this mode, the ADC uses the conversion clock as the ADC clock source both when active or idle. To minimize conversion latency, it is recommended the conversion clock be configured to 5.0 MHz. No startup delay (defined by PUDELAY in the POWER register) is imposed.
2. Auto power-down mode
This mode operates when:
— At least one ADC converter is powered up (PD0 or PD1=0 in the POWER register);
— Auto power-down mode is enabled (APD=1 in the POWER register);
— The ADC’s clock is enabled (ADC=1 in the SIM module’s SIM_PCE register).
Auto power-down and standby modes can be used together by setting APD=1 in the above configuration.
This hybrid mode converts at an ADC clock rate of 100 kHz using standby current mode when active, and gates off the ADC clock and powers down the converters when idle. A startup delay of PUDELAY ADC clock cycles execute at the start of all scans while the ADC engages the conversion clock and the ADC powers up, stabilizing in the standby current mode. This provides the lowest possible power configuration for ADC operation.
3. Auto standby mode
This mode operates when:
— At least one ADC converter is powered up (PD0 or PD1=0 in the POWER register);
— Auto power-down is disabled (APD=0 in the POWER register);
— Auto standby is enabled (ASB=1 in the POWER register);
28-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
— The ADC’s clock is enabled (ADC=1 in the SIM module’s SIM_PCE register);
— Either the relaxation oscillator must be enabled for 8-MHz operation or the external oscillator clock must be running at 8 MHz in this mode.
In auto standby mode, the ADC uses the conversion clock when active and the100 kHz Standby clock when idle. The standby (low current) state automatically engages when the ADC is idle. The ADC will execute a startup delay of PUDELAY ADC clocks at the start of all scans, allowing the ADC to switch to the Conversion clock and to revert from standby to normal current mode.
It is recommended the conversion clock be configured at or near 5.0 MHz to minimize conversion latency when active. In this mode, the ADC uses the conversion clock when active and gates off the conversion clock and powers down the converters when idle. A startup delay of PUDELAY ADC clocks is executed at the start of all scans, allowing the ADC to stabilize when switching to normal current mode from a completely powered off condition. This mode uses less power than normal and more power than auto standby. It requires more startup latency than auto standby when leaving the idle state to start a scan (higher
PUDELAY value).
4. POWER-DOWN MODE
This mode operates when:
— Both ADC converters are powered down (PD0=PD1=1 in the POWER register);
— The ADC’s clock is disabled (ADC=0 in the SIM module’s SIM_PCE register).
In this configuration, the clock trees to the ADC and all of its analog components are shut down and the ADC uses no power.
28.4.6.2
Power Management Details
The ADC voltage reference and converters are powered down (PDn=1 in the POWER register) on reset.
Individual converters can be manually powered down when not in use (PD0=1 or PD1=1), and the voltage reference can be automatically powered down when no converter is in use (PD2=1) or manually powered up when no converters are powered (PD2=0). When the ADC voltage reference is powered down, output reference voltages are set to low (V
SSA
).
A delay of PUDELAY ADC clock cycles is imposed whenever PD0 or PD1 are cleared to power-up a converter and whenever the ADC goes from an idle (neither converter has a scan in process) to an active state when not operating in normal power mode. The ADC is active when at least one converter has a scan in process. A device recommends the use of two PUDELAY values: a large value for full power-up and a smaller value for going from standby current levels to full power-up. The following paragraphs provide an explanation of how to use PUDELAY when starting the ADC up or changing modes.
When starting up in normal mode, first set PUDELAY to the large power-up value. Next, clear the PD0 and or PD1 bits to power-up the required converters. Poll the status bits (PSTS n in the POWER register) until all required converters are powered up. Following polling, start scan operations. The value in
PUDELAY will provide a power-up delay before scans begin. Since normal mode does not use PUDELAY at start of scans, no further delays will be imposed.
When starting up using auto standby mode, first use the normal mode startup procedure. Before starting scan operations, set PUDELAY to the smaller value, then set ASB in the POWER register. Auto standby
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-13
Analog-to-Digital Converter (ADC) mode will automatically reduce current levels until active and then impose a PUDELAY wait to allow current levels to rise from standby to normal levels.
When starting up using auto power-down mode, first use the normal mode startup procedure. Before starting scan operations, set PUDELAY to the large power-up value. Next, set APD in the POWER register. Finally, clear the PD0 and or PD1 bits for the required converters. Converters remain powered off until scanning goes active, at which time the large PUDELAY executes as the ADC goes from powered down to fully powered at the start of the scan.
In auto power-down mode, when the ADC goes from idle to active, a converter is only powered up if it is required for the scan, as determined by the ADLST1, ADLST2, and SDIS registers.
It is recommended to power-off both converters (PD0=PD1=1 in the POWER register) when re-configuring clocking or power controls to avoid generating bad samples and ensure proper delays are applied when powering up or starting scans.
Attempts to start a scan during the PUDELAY time-out will be ignored until the appropriate PSTS n bits are cleared in the POWER register.
Any attempt to use a converter when powered down or with the voltage reference disabled results in invalid results. It is possible to read ADC result registers after converter power down to see results calculated before power-down. However, a new scan sequence must be started with a SYNC n pulse or a write to the START n bit before new results will be available.
28.4.6.3
ADC STOP Mode of Operation
Any conversion sequence in progress can be stopped by setting the relevant STOP n bit. Any further sync pulses or writes to the START n bit are ignored until the STOP n bit is cleared. Once in this stop mode, the results registers can be modified by writes from the processor. Any write to RSLT n registers in the ADC stop mode is treated as if the analog core supplied the data, so limit checking, zero crossing, and associated interrupts can occur if enabled.
28.4.7
ADC Clock
28.4.7.1
General
The ADC has two external clock inputs used to drive two clock domains within the ADC module.
Table 28-3. ADC Clock Summary
Clock input Source Characteristics
Peripheral Clock
(=System Clock)
1/2 Core clock Maximum rate is PLL output divided by 2 if PLL enabled. When PLL disabled, max rate is oscillator clock divided by 2.
ADC 8MHz Clock Relaxation
Oscillator (8Mhz),
Crystal Oscillator
(1-16MHz), or external Oscillator
Provides 8Mhz for auto standby power saving mode.
28-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
28.4.7.2
Description of Clock Operation
the conversion clock is the primary source for the ADC clock and is always selected as the ADC clock when conversions are in process. The DIV value in the CTRL2 register should be configured so the conversion clock frequency falls between 100 kHz and 5.0 MHz. Operating the ADC at out-of-spec clock frequencies will degrade conversion accuracy. Similarly, modifying the parameters affect clock rates or power modes while the regulators are powered up (PD0=0 or PD1=0) will also degrade conversion accuracy.
The conversion clock ADC uses for sampling is calculated using the IPBus clock and the clock divisor bits
within the CTRL2 register. Please see Section 28.5.1 “Control 1 Register (CTRL1) ” or
“Control 2 Register (CTRL2) Under Sequential Scan Modes
” .
The ADC clock is active 100% of the time while in loop modes, or if power management is set to normal. It is also active during all ADC power-up for a period of time determined by the PUDELAY field in the power (POWER) register. After the power-up delay times out, the ADC clock continues until the completion of the ADC n scan when operating in auto standby or auto power-down modes.
Auto Standby
Enabled (ASB=1)
Auto Power-Down
Disabled (APD=0)
ADC Idle
Standby Current Mode Status
CTRL2:DIV
0
Peripheral (System) Clock
1/2 Core Frequency
Oscillator Clock
(8MHz)
(+80)
0
1
DIV
(+2 x [DIV+1])
Auto Standby Clock
(100kHz)
ADC
Conversion
Clock
0
1
ADC Bit in SIM_PCE
ADC Clock
Figure 28-8. ADC Clock Generation
The oscillator clock feeds an 80:1 divider, generating the auto standby clock. The auto standby clock is selected as the ADC clock during the auto standby power mode when both converters are idle. The auto standby power mode requires an 8 MHz oscillator clock from the relaxation oscillator, crystal oscillator, or external oscillator.
28.4.7.3
ADC Clock Resynchronization at Start of Scan
At the fastest ADC speed, each ADC clock period is 6 system clock periods long. When asserting the start of a scan, either by writing to a START n bit or by a SYNC n signal, the ADC clock is re-synchronized to align it to the system clock. This allows the commanded scan to begin as soon as possible rather than wait
for both sequential and simultaneous parallel modes of operation. In these modes, both ADC operate off of the same start signal.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-15
Analog-to-Digital Converter (ADC)
In a parallel scan mode when SIMULT=0, both ADCs operate using independent START n bits and
SYNC n signals. As shown in
, the first scan started will be re-synchronized to the system clock, but the second scan may wait up to 5 additional system clocks before starting. Also, note that which converter is synchronized to the system clock depends on which convert first starts to use the ADC. The case shown has ADCA synchronized, but one could easily imagine the case where the ADCA start comes after instead of before the ADCB start. In this case, ADCAs start would be delayed up to 5 additional system clock periods instead of ADCBs.
If there is a known timing relationship between ADCA and ADCB when operating in a non-simultaneous parallel mode, then the application can control which ADC starts first and gets the re-synchronized clock.
The application can also control the delay to starting the second ADC scan so that its start signal aligns with the ADC clock, and the start of the second ADC is not delayed.
START0
Asserted
ADC Conversion Clock Resynchronized
ADC Scans Start
System Clock
Old ADC Clock
ADC Clock After
Resynchronization
ADCA Scan
ADCB Scan
Figure 28-9. ADC Clock Resynchronization for
Sequential and Simultaneous Parallel Modes
28-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
ADC Conversion Clock Resynchronized
ADCA Scan Start
START1 Asserted
ADCB Scan Should Start Here
ADCB Scan Start
System Clock
START0
Asserted
Old ADC Clock
Wait for next rising edge of ADC
Conversion Clock
ADC Clock After
Resynchronization
ADCA Scan
ADCB Scan
Delay in start because ADC Clock cannot be resynchronized: 5 System Clocks
Figure 28-10. ADC Clock Resynchronization for
Non-Simultaneous Parallel Modes
28.4.8
Voltage Reference Pins V
REFH
& V
REFL
The voltage difference between V
REFH
and V
REFL
provides the reference voltage that all analog inputs are measured against. The reference voltage should be provided from a low noise filtered source capable of providing up to 1mA of reference current.
1.0mH
0.1µF
External
Reference
Voltage
V
REFH
V
RH
V
RL
SEL_VREFH
V
REFH
to ADC
V
REFL
to ADC
SEL_VREFL
V
REFL
Figure 28-11. ADC Voltage Reference Circuit
When tying V
REFH
to the same potential as V the amplitude of V
DDA
DDA,
relative measurements are being made with respect to
. It is imperative that special precautions be taken to assure the voltage applied to
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-17
Analog-to-Digital Converter (ADC)
V
REFH
is as noise-free as possible. Any noise residing on the V
REFH
voltage is directly transferred to the digital result.
Figure 28-11 illustrates the internal workings of the ADC voltage reference circuit. V
REFH filtered; a minimum configuration is shown in the figure.
must be noise
28.4.9
Supply Pins V
DDA
and V
SSA
Dedicated power supply pins are provided for the purposes of reducing noise coupling and to improve accuracy. The power provided to these pins is suggested to come from a low noise filtered source.
Uncoupling capacitors ought to be connected between V
DDA
and V
SSA
.
28.5
Register Definitions
A register address is the sum of a base address and an address offset. The base address is defined at the device level, and the address offset is defined at the module level.
lists the ADC registers in ascending address order, including their acronyms and address offset of each register. ADC uses ADC n _BASE plus the given offset depending on the ADC being used. The
ADC peripheral has 43 registers.
Table 28-4. ADC Memory Map
IPSBAR Offset
1
Register
Width
(bits)
Access Reset Value Section/Page
Supervisor Mode Access Only
0x0019_0000 Control Register 1 (CTRL1)
0x0019_0002 Control Register 2 (CTRL2)
0x0019_0004 Zero Crossing Control Register (ADZCC)
0x0019_0006 Channel List Register 1 (ADLST1)
0x0019_0008 Channel List Register 2 (ADLST2)
0x0019_000A Sample Disable Register (ADSDIS)
0x0019_000C Status Register (ADSTAT)
0x0019_000E Limit Status Register (ADLSTAT)
16
16
16
16
16
16
16
16
0x0019_0010 Zero Crossing Status Register (ADZCSTAT)
0x0019_0012–20 Result Registers 0-7 (ADRSLT0-7)
0x0019_0022–30 Low Limit Registers 0-7 ()
0x0019_0032–40 High Limit Registers 0-7 (ADLLMT0-7)
16
16
16
16
0x0019_0042–50 Offset Registers 0-7 (ADOFS0-7)
0x0019_0052 Power Control Register (POWER)
16
16
R/W
R/W
0x0019_0054 Voltage Reference Register (CAL) 16 R/W
1
Addresses not assigned to a register and undefined register bits are reserved for expansion.
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
0x5005
0x0002
0x0000
0x3210
0x7654
0x0000
0x0000
0x0000
0x0000
0x0000
0x0000
0x0000
0x0000
0x00D7
0x0000
28-18
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
IPSBAR Offset
0x0019_0000
0x0019_0002
Acronym
CTRL1
CTRL2
Table 28-5. ADC Register Summary
Register Name
Control Register 1
Control Register 2
Access Type
Read/Write
Read/Write
ADZCC
ADLST1
ADLST2
ADSDIS
ADSTAT
ADLSTAT
ADZCSTAT
ADRSLT0-7
ADLLMT0-7
ADHLMT0-7
ADOFS0-7
POWER
CAL
Zero Crossing Control Register
Channel List Register 1
Channel List Register 2
Sample Disable Register
Status Register
Limit Status Register
Zero Crossing Status Register
Result Registers 0-7
Low Limit Registers 0-7
High Limit Registers 0-7
Offset Registers 0-7
Power Control Register
Voltage Reference Register
Location
0x0019_0004
0x0019_0006
0x0019_0008
0x0019_000A
0x0019_000C
0x0019_000E
0x0019_0010
0x0019_0012–20
0x0019_0022–30
0x0019_0032–40
0x0019_0042–50
0x0019_0052
0x0019_0054
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Bits of each of the 43 registers are summarized in
Details of each follow.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-19
Analog-to-Digital Converter (ADC)
Add. Offset
$0019_0000
$0019_0002
$0019_0004
$0019_0006
$0019_0008
$0019_000A
Register
Acronym
CTRL 1
CTRL 2
Simultaneous Mode
R
W
R
CTRL 2
Parallel Mode
W
ADZCC
ADLST1
ADLST2
ADSDIS
W
R
R
W
R
W
R
W
R
W
15
0
0
0
0
0
0
14 13 12 11 10 9 8
0
STOP0
START
0
SYNC
0
EOSIE
0
ZCIE
LLMTI
E
HLMTI
E
0 0 0 0 0 0 0
0
STOP1
START
1
SYNC
1
EOSIE
1
ZCE7
0
ZCE6
SAMPLE3
SAMPLE7
0 0
0
0
0
ZCE5
0
0
0
SAMPLE2
SAMPLE6
0
ZCE4
0
0
7
0
0
0
0
ZCE3
6 5
CHNCFG
0
0
0
SIMULT
ZCE2
SAMPLE1
SAMPLE5
4 3
0
0
0
ZCE1
2
DIV
DIV
1
SMODE
ZCE0
SAMPLE0
SAMPLE4
0
DS7 DS6 DS5 DS4 DS3 DS2 DS1 DS0
Figure 28-12. ADC Register Map Summary
28-20
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
R CIP0 CIP1 0
EOSI
1
EOSI
0
ZCI LLMTI HLMT RDY7 RDY6 RDY5 RDY4 RDY3 RDY2 RDY1 RDY0
$0019_0006 ADSTAT
$0019_0007
$0019_0008
$0019_0009-10
ADLSTAT
ADZCSTAT
ADRSLT n
$0019_0011-$18 ADLLMT n
W
R
W
R
W
HLS7 HLS6 HLS5 HLS4 HLS3 HLS2 HLS1 HLS0 LLS7 LLS6 LLS5 LLS4 LLS3 LLS2 LLS1 LLS0
R 0 0 0 0 0 0 0 0
ZCS7 ZCS6 ZCS5 ZCS4 ZCS3 ZCS2 ZCS1 ZCS0
W
R
SEXT
W
0
RSLT
TEST_DATA
LLMT
0
0
0
0
0
0
$0019_0019-$20 ADHLMT n
R
W
0
HLMT
0 0 0
$0019_0021-$28 ADOFS n
R
W
0
OFFSET
0 0 0
$0019_0029 POWER
R
ASB
0 0
PSTS
2
PSTS
1
PSTS
0
PUDELAY APD PD2 PD1 PD0
W
0 0 0 0 0 0 0 0 0 0 0 0 0 0
$0019_002A CAL
R
W
SEL_V
REFH
SEL_V
REFL
R
W
0 Read as 0
Reserved
Figure 28-12. ADC Register Map Summary (continued)
28.5.1
Control 1 Register (CTRL1)
Bits 14, 13, 12, and 11 in CTRL1 control all types of scans except parallel scans in the B converter when
SIMULT=0 in the CTRL2 register. SIMULT=0 bits 14, 13, 12, and 11 in CTRL are used to control converter B scans in parallel scan modes, while the equivalent bits in CTRL1 are used for converter A.
Access: read/write IPSBAR
Offset:
0x19_0000 (CTRL1)
15
R 0
W
Reset 0
14 13 12 11
STOP0
START0
SYNC0
EOSI
E0
1 0 1 0
10
ZCIE
9 8
LLMT
IE
HLMT
IE
0 0 0
7
0
6
0
Figure 28-13. Control 1 (CTRL1) Register
5
CHNCFG
0
4
0
3
0
2 1
SMODE
0
1 0 1
28.5.1.1
Reserved—Bit 15
This bit is reserved or not implemented. It is read as 0 and cannot be modified by writing.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-21
Analog-to-Digital Converter (ADC)
28.5.1.2
STOP 0 (STOP0)—Bit 14
When STOP0 is asserted, the current scan is stopped and no further scans can start. Any further SYNC0 input pulses (see SYNC0 bit 12) or writes to the START0 bit are ignored until the STOP0 bit is cleared.
After the ADC is in stop mode, the result registers can be modified by the processor. Any changes to the result registers in stop mode are treated as if the analog core supplied the data. Therefore, limit checking, zero crossing, and associated interrupts can occur if enabled.
NOTE
This is not the same as the device’s STOP mode.
• 0 = Normal operation
• 1 = Stop mode
28.5.1.3
Start Conversion (START0)—Bit 13
A scan is started by writing 1 to the START0 bit. This is a write-only bit. Writing 1 to the START0 bit again will be ignored until the end of the current scan.
• 0 = No action
• 1 = Start command is issued
The ADC must be in a stable power configuration prior to writing the START bit. Refer to the functional description of power modes for further details.
28.5.1.4
Synchronization 0 Enable (SYNC0)—Bit 12
A conversion may be initiated by asserting a positive edge on the SYNC0 input. Any subsequent SYNC0 input pulses while the scan remains in process are ignored.
• 0 = Scan is initiated by a write to START0 bit only
• 1 = Use a SYNC0 input pulse or START0 bit to initiate a scan
The ADC must be in a stable power mode prior to SYNC0 input assertion. Refer to the functional description of power modes for further details.
In “ OnCE Sequential” and “ OnCE Parallel” scan modes, only the first SYNC0 input pulse is honored.
Subsequent SYNC0 input pulses are ignored until SYNC0 input is re-armed by writing to the CTRL1 register, usually by simply rewriting 1 to SYNC0. This is achieved at any time, even during the execution of the scan.
28.5.1.5
End Of Scan Interrupt Enable 0 (EOSIE0)—Bit 11
This bit enables an EOSI0 interrupt to be generated upon completion of the scan. For looping scan modes, the interrupt will trigger after the completion of each iteration of the loop.
• 0 = Interrupt disabled
• 1 = Interrupt enabled
28-22
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
28.5.1.6
Zero Crossing Interrupt Enable (ZCIE)—Bit 10
This bit enables the zero crossing interrupt if the current result value has a sign change from the previous result as configured by the ADZCC register.
• 0 = Interrupt disabled
• 1 = Interrupt enabled
28.5.1.7
Low Limit Interrupt Enable (LLMTIE)—Bit 9
This bit enables the low limit exceeded interrupt when the current result value is less than the low limit register value. The raw result value is compared to the LOLIM register, bits LLMT[11:0], before the offset register value is subtracted.
• 0 = Interrupt disabled
• 1 = Interrupt enabled
28.5.1.8
High Limit Interrupt Enable (HLMTIE)—Bit 8
This bit enables the high limit exceeded interrupt if the current result value is greater than the high limit register value. The raw result value is compared to the high limit (HILIM) register, bits HLMT[11:0], before the offset register value is subtracted.
• 0 = Interrupt disabled
• 1 = Interrupt enabled
28.5.1.9
Channel Configure (CHNCFG)—Bits 7 – 4
The inputs can be configured for either single-ended or differential conversions.
Table 28-6. CHNCFG Bit Settings
Bit Settings xxx1 xxx0 xx1x xx0x x1xx x0xx
1xxx
0xxx
Inputs
AN0 – AN1
AN2 – AN3
AN4 – AN5
AN6 – AN7
Description
Configured as differential pair (AN0 is + and AN1 is – )
Both configured as single ended inputs
Configured as differential pair (AN2 is + and AN3 is – )
Both configured as single ended inputs
Configured as differential pair (AN4 is + and AN5 is – )
Both configured as single ended inputs
Configured as differential pair (AN6 is + and AN7 is – )
Both configured as single ended inputs
Differential measurements return the max value 32760 (= 4095 × 8) when the plus ( + ) input is V the minus ( −) input is V
REFL
, return 0 when the plus ( + ) input is at V
REFL
V
REFL measurements return the max value 32760 when the input is at V
REFH
REFH
and
and the minus ( −) input is at
, and scale linearly between based on the voltage difference between the two signals. Single-ended
, return 0 when the input is at V and scale linearly between based on the amount by which the input exceeds V
REFL
.
REFL
,
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-23
Analog-to-Digital Converter (ADC)
28.5.1.10 Scan Mode Control (SMODE)—Bits 2-0
SMODE controls the scan mode of the ADC module. All scan modes make use of the 8 sample slots defined by the ADLST1 and ADLST2 registers. A scan is the process of stepping through these sample slots, converting the analog input indicated by that slot, and storing the result. Slots that are not required may be disabled by writing 1 to the appropriate bits of the SDIS register.
Input pairs AN0-1, AN2-3, AN4-5, and AN6-7 may be configured as differential pairs using the CHNCFG field. When a slot in CLST n refers to either member of a differential pair, a differential measurement on that pair will be made; otherwise, a single-ended measurement will be taken on that input. The details of differential and single-ended measurement are described in the description of the CHNCFG field.
SMODE determines whether the slots are used to perform a sequential scan of up to 8 samples or 2 parallel scans up to 4 samples. SMODE controls how these scans are initiated and terminated. It also controls whether the scans are performed once or repetitively. For more details, please see
“Parallel Mode Operation of the ADC
” .
Parallel scans may be simultaneous (SIMULT=1) or non-simultaneous (SIMULT=0). During simultaneous parallel scans, A and B converters scan synchronously using one set of shared controls
(CTRL1 register). During non-simultaneous scans, the A and B converters operate asynchronously with each converter using its own independent set of controls (CTRL1 for A and CTRL2 for B). Refer to the
SIMULT bit description for further details.
NOTE
The SIMULT bit only applies to parallel operating modes and is ignored during sequential operating modes.
• 000 = Once sequential
Upon START or an enabled sync signal, samples are taken one at a time starting with SAMPLE0 until a first disabled sample is encountered. If no disabled sample is encountered in SDIS register, conversion concludes after SAMPLE7. If the scan is initiated by a sync signal, only one scan will be completed until the converter is rearmed by writing to the CTRL1 register.
• 001 = Once parallel
Upon START or an armed and enabled sync signal, converter A will capture samples 0-3 and converter B will capture samples 4-7. By default (SIMULT=1), samples are taken simultaneously
(synchronously), and scanning stops when either converter encounters a disabled sample or both converters complete all 4 samples. When SIMULT=0, samples are taken asynchronously, and scanning stops when each converter encounters a disabled sample in its part of the SDIS register or completes all 4 samples. If the scan is initiated by a sync signal, only one scan will be completed till the converter is re-armed by writing to the CTRL1 register. (When SIMULT=0, the B converter must be re-armed separately by writing to the CTRL2 register.)
• 010 = Loop sequential
Upon an initial start or enabled sync pulse, up to 8 samples are taken one at a time until a disabled sample is encountered. The process repeats until the STOP0 bit is set. While a loop mode is running, any additional start commands or sync pulses are ignored. If auto standby
(POWER:ASB=1) or auto power-down (POWER:APD=1) is the selected power mode control, the power-up delay defined by PUDELAY will be applied only on the first conversion.
28-24
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
• 011= Loop parallel
Upon an initial start or enabled sync pulse, converter A will capture Samples 0-3, and converter B will capture Samples 4-7. Each time a converter completes its current scan, it immediately restarts its scan sequence. This continues until a STOP n bit is asserted. While a loop is running, any additional start commands or sync pulses are ignored. By default (SIMULT=1), samples are taken simultaneously (synchronously), and scanning stops when either converter encounters a disabled sample or both converters complete all 4 samples. When SIMULT=0, samples are taken asynchronously, and scanning stops when each converter encounters a disabled sample in its part of the SDIS register or completes all 4 samples. If auto standby or auto power-down is the selected power mode control, the power-up delay defined by PUDELAY will be applied only on the first conversion.
• 100 = Triggered sequential
Upon START or an enabled sync signal, samples are taken one at a time starting with SAMPLE0 until a first disabled sample is encountered. If no disabled sample is encountered, conversion concludes after SAMPLE7. If external sync is enabled, new scans will be started for each sync pulse that is non-overlapping with a current scan in progress.
• 101 = Triggered parallel (default)
Upon START or an enabled sync signal, converter A will convert Samples 0-3, and converter B will convert Samples 4-7 in parallel. By default (SIMULT=1), samples are taken simultaneously
(synchronously), and scanning stops when either converter encounters a disabled sample or both converters complete all 4 samples. When SIMULT=0, samples are taken asynchronously, and scanning stops when each converter encounters a disabled sample in its part of the SDIS register or completes all 4 samples. If external sync is enabled (SYNC0=1), new scans will be started for each sync pulse as long as the ADC has completed the previous scan (STAT:CIP n =0).
• 110 = Reserved use
• 111 = Reserved use
28.5.2
Control 2 Register (CTRL2) Under Sequential Scan Modes
Operating mode dependencies occur when the ADC’s scan mode (SMODE in the CTRL1 register) is set to once sequential, loop sequential, or triggered sequential. Bits 15
–
5 are reserved. Only the DIV field is available.
IPSBAR
Offset:
0x19_0001 (CTRL2)
15
R 0
W
Reset 0
14
0
13
0
12
0
11
0
10
0
9
0
8
0
7
0
6
0
5
0
4 3
0 0 0 0 0 0 0 0 0 0 0 0
Figure 28-14. Control 2 (CTRL2) Register Under Sequential Scan Modes
Access: read/write
2
DIV
0
1
1
0
0
28.5.2.1
Reserved—Bits 15–5
This bit field is reserved and should not be modified by writing.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-25
Analog-to-Digital Converter (ADC)
28.5.2.2
Clock Divisor Select (DIV)—Bits 4 – 0
The divider circuit generates the ADC clock by dividing the system clock by 2 × (DIV[4:0]+1). A DIV value must be chosen so the ADC clock does not exceed 5.0 MHz. The following table shows ADC clock frequency based on the value of DIV for several configurations.
Table 28-7. ADC Clock Frequency for Various Conversion Clock Sources
DIV
0_0000
0_0001
0_0010
0_0011
0_0100
—
—
1_1111
Divisor
10
—
—
64
6
8
2
4
ROSC Standby
400 Khz
200 kHz Sys Clock
100K
100K
100K
100K
100K
—
—
100K
ROSC Normal
8 Mhz
4 MHz Sys Clock
2.00M
1.00M
500K
250K
125K
—
—
62.5K
PLL
64 Mhz
32 MHz Sys Clock
16.0M
8.00M
5.33M
4.00M
3.20M
—
—
500K
External CLK
CLK/2 Sys Clock
CLK/4
CLK/8
CLK/12
CLK/16
CLK/20
—
—
CLK/128
28.5.3
Control 2 Register (CTRL2) Under Parallel Scan Modes
Operating mode dependencies of this register occur when the ADC’s scan mode (SMODE in the CTRL1 register) is set to once parallel, loop parallel, or triggered parallel. Bits 14
–
11 and 5 are no longer reserved and are used to control the operation of converter B.
By default, SIMULT=1 and converter B operates together with converter A. In this case bits, 14, 13, 12, and 11 in CTRL2 do not affect converter B operation. When SIMULT=0 and SMODE is a parallel scan, bits 14
–
11 in CTRL2 along with the SYNC1 input are used to control the converter B scan. In this case,
EOSIE1 enables the EOSI1 interrupt, signaling the end of a B converter scan. Also, the CIP1 bit in the
STAT register is used to indicate a converter B scan is active.
Access: read/write IPSBAR
Offset:
0x19_0001 (CTRL2)
15
R 0
W
Reset 0
14 13 12 11
STOP1
1
START1
SYNC1 EOSIE1
0 1 0
10
0
0
9
0
0
8
0
0
7
0
0
6
0
0
5
SIMU
LT
0
4
0
3
0
Figure 28-15. Control 2 (CTRL2) Register Under Parallel Scan Modes(
2
DIV
0
1
1
0
0
28.5.3.1
Reserved—Bit 15
This bit is reserved or not implemented. It is read as 0 and cannot be modified by writing.
28-26
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
28.5.3.2
Stop (STOP1)—Bit 14
During parallel scan modes when SIMULT=0, setting STOP1 stops parallel scans in the B converter and prevents new ones from starting. Any further SYNC1 input pulses (please see SYNC1 bit) or writes to the
START1 bit are ignored until the STOP1 bit is cleared. After the ADC is in Stop mode, the B converter results registers can be modified by the processor. Any changes to the Result registers in Stop mode are treated as if the analog core supplied the data. Therefore, limit checking, zero crossing, and associated interrupts can occur if enabled. This is not the same as the device’s STOP mode .
• 0 = Normal operation
• 1 = Stop command issued
28.5.3.3
Start Conversion (START1)—Bit 13
During parallel scan modes when SIMULT=0, a B converter parallel scan is started by writing 1 to the
START1 bit. This is a write-only bit. Writing 1 to the START1 bit again will be ignored until the end of the current scan.
• 0 = No action
• 1 = Start a B converter parallel scan
The ADC must be in a stable power configuration prior to writing the start bit. Refer to the functional description of power modes for further details.
28.5.3.4
SYNC1 Enable (SYNC1)—Bit 12
During parallel scan modes when SIMULT=0, setting SYNC1 to 1 permits a B converter parallel scan to be started by asserting the SYNC1 input for at least one ADC clock cycle. Any additional SYNC1 input pulses will be ignored until the end of the scan.
• 0 = B converter parallel scan is initiated by a write to START1 bit only
• 1 = Use a SYNC1 input pulse or START1 bit to initiate a B converter parallel scan
The ADC must be in a stable power mode prior to SYNC1 input assertion. Please refer to the functional description of power modes for further details.
In “ Once Sequential” and “ Once Parallel” scan modes, only a first SYNC1 input pulse is honored.
Subsequent SYNC1 input pulses are ignored until the SYNC1 input is re-armed by writing to the CTRL2
Register, usually by simply rewriting 1 to SYNC1. This can be done at any time, including while the scan remains in process.
28.5.3.5
Reserved—Bits 10–6
This bit field is reserved or not implemented. It is read as 0 and cannot be modified by writing.
28.5.3.6
End Of Scan Interrupt Enable 1 (EOSIE1)—Bit 11
During parallel scan modes when SIMULT=0, this bit enables an EOSI1 interrupt to be generated upon completion of a B converter parallel scan. For looping scan mode, the interrupt will trigger upon the completion of each iteration of the loop .
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-27
Analog-to-Digital Converter (ADC)
• 0 = Interrupt disabled
• 1 = Interrupt enabled
28.5.3.7
Simultaneous Mode (SIMULT)—Bit 5
This bit only affects parallel scan modes.
When SIMULT=1 (default value) parallel scans operate in simultaneous mode. The scans in the A and B converter operate simultaneously and always result in pairs of simultaneous conversions in the A and B converter. START0, STOP0, SYNC0, and EOSIE0 control bits and the SYNC0 input are used to start and stop scans in both converters simultaneously. A scan ends in both converters when either converter encounters a disabled sample slot. When the parallel scan completes, the EOSI0 triggers if EOSIE0 is set.
The CIP0 status bit indicates that a parallel scan is in process.
When SIMULT=0, parallel scans in the A and B converters operate independently. The B converter has its own independent set of the above controls (START1, STOP1, SYNC1, EOSIE1, SYNC1) designed to control its operation and report its status. Each converter
’ s scan continues until its sample list is exhausted
(four samples) or a disabled sample is encountered. For looping parallel scan mode, each converter starts its next iteration as soon as the previous iteration in that converter is complete and continues until the
STOP bit for that converter is asserted.
• 0 = Parallel scans done independently
• 1 = Parallel scans done simultaneously (default)
28.5.3.8
Clock Divisor Select (DIV)—Bits 4 – 0
The divider circuit generates the ADC clock by dividing the system clock by 2 × (DIV[4:0]+1). A DIV
value must be chosen so the ADC clock does not exceed 5.0Mhz. Table 28-8 shows ADC clock frequency
based on the value of DIV for various configurations.
Table 28-8. ADC Clock Frequency for Various Conversion Clock Sources
DIV
0_0000
0_0001
0_0010
0_0011
0_0100
—
—
1_1111
Divisor
10
—
—
64
6
8
2
4
ROSC Standby
400Khz
200kHz Sys Clock
100K
100K
100K
100K
100K
—
—
100K
ROSC Normal
8Mhz
4MHz Sys Clock
2.00M
1.00M
500K
250K
125K
—
—
62.5K
PLL
64 Mhz
32MHz Sys Clock
16.0M
8.00M
5.33M
4.00M
3.20M
—
—
500K
External CLK
CLK/2 Sys Clock
CLK/4
CLK/8
CLK/12
CLK/16
CLK/20
—
—
CLK/128
28-28
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
28.5.4
Zero Crossing Control Register (ADZCC)
The ADC zero crossing control (ADZCC) register provides the ability to monitor the selected channels and determine the direction of zero crossing triggering the optional interrupt. Zero crossing logic monitors only the sign change between current and previous sample. The ZCE0 bit monitors the sample stored in
ADRSLT0, ZCE1 bit monitors ADRSLT1, and ZCE7 bit monitors ADRSLT7. When the zero crossing is disabled for a selected result register, sign changes are not monitored or updated in the ADZCSTAT register.
Access: read/write IPSBAR
Offset:
0x19_0002 (ADZCC)
13 15 14
R
W
Reset 0
ZCE7
0
ZCE6
12 11
ZCE5
10 9
ZCE4
8 7
ZCE3
6 5
ZCE2
4
0 0 0 0 0 0 0 0 0 0
Figure 28-16. Zero Crossing Control (ADZCC) Register
3
ZCE1
2
0 0
1
ZCE0
0
0 0
28.5.4.1
Zero Crossing Enable n ( ZCE n )—Bits 15–0
For each channel, n , setting the ZCE n field allows detection of the indicated zero crossing condition, provided the corresponding offset register (ADOFS n ) has a value offset , 0 < offset < 0x7FF8.
• 00 = Zero crossing disabled
• 01 = Zero crossing enabled for positive to negative sign change
• 10 = Zero crossing enabled for negative to positive sign change
• 11 = Zero crossing enabled for any sign change
28.5.5
Channel List 1 and 2 Registers (ADLST1 and ADLST2)
The channel list register contains an ordered list of the analog input channels to be converted when the next scan is initiated. If all samples are enabled in the SDIS register, a sequential scan of inputs proceeds in order of SAMPLE0 through SAMPLE7. If one of the parallel sampling modes is selected instead, the converter A sampling order is SAMPLE0-3, and the converter B sampling order is SAMPLE4-7.
Access: read/write IPSBAR
Offset:
0x19_0003 (ADLST1)
14 13 15
R
W
Reset 0 0
SAMPLE3
1
12 11 10 9
SAMPLE2
8 7 6 5
SAMPLE1
4
1 0 0 1 0 0 0 0
Figure 28-17. Channel List 1 (ADLST1) Register
1
3
0
2 1
SAMPLE0
0
0 0 0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-29
Analog-to-Digital Converter (ADC)
IPSBAR
Offset:
0x19_0004 (ADLST2)
14 13 15
R
W
Reset 0 1
SAMPLE7
1
12 11 10 9
SAMPLE6
8 7 6 5
SAMPLE5
4
1 0 1 1 0 0 1 0
Figure 28-18. Channel List 2 (ADLST2) Register
1
3
0
Access: read/write
2 1
SAMPLE4
0
1 0 0
28.5.5.1
Reserved—Bits 15, 11, 7 and 3
These bits are reserved or are not implemented. They are read as 0 and cannot be modified by writing.
28.5.5.2
SAMPLE n (SAMPLE4)—Bits 2, 1, and 0
The value of the SAMPLE n field is used to select the input channel to be sampled.
Table 28-9. ADC Input Conversion for Sample Bits
SAMPLE n [2:0]
ADC Input Pins Selected
Sequential Mode n =0,1,2,...,7
000
001
010
011
100
101
110
111 n =0,1,2,3
(Conv. A)
000
001
010
011
Parallel Mode n =4,5,6,7
(Conv. B)
100
101
110
111
Single Ended
AN0
AN1
AN2
AN3
AN4
AN5
AN6
AN7
Differential
AN0+, AN1 –
AN2+, AN3 –
AN4+, AN5 –
AN6+, AN7 –
In sequential modes, the sample slots are converted in order from SAMPLE0 to SAMPLE7. Analog input pins can be sampled in any order, including sampling the same input pin more than once.
In parallel modes, converter A processes sample slots SAMPLE0 through SAMPLE3, while converter B processes sample slots SAMPLE4 through SAMPLE7. Since converter A only has access to analog inputs
AN0 through AN3, sample slots SAMPLE0-3 should only contain binary values between 000 and 011.
Likewise, since converter B only has access to analog inputs AN4 through AN7, sample slots SAMPLE4-7 should only contain binary values between 100 and 111. No damage will occur if this constraint is violated but results are undefined.
28-30
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
When inputs are configured as differential pairs, a reference to either analog input in a differential pair by a sample slot implies a differential measurement on the pair. The details of single ended and differential measurement are described under the CHNCFG field. Sample slots are disabled using the SDIS register.
28.5.6
Sample Disable Register (SDIS)
This register is an extension to the ADLST1and ADLST2, providing the ability to enable only the desired samples programmed in the SAMPLE0–SAMPLE7. At reset, all samples are enabled. For example, if in sequential mode and bit DS5 is set to 1, SAMPLE0 through SAMPLE4 are sampled. However, if in parallel mode and bits DS5 or DS1 are set to 1, only SAMPLE0 and SAMPLE4 are sampled.
Access: read/write IPSBAR
Offset:
0x19_0005 (SDIS)
15
R 0
W
Reset 0
14
0
0
13
0
0
12
0
0
11
0
0
10
0
0
9
0
8
0
7 6 5 4 3 2 1 0
DS7 DS6 DS5 DS4 DS3 DS2 DS1 DS0
0 0
Figure 28-19.
0 0 0 0 0 0 0 0
28.5.6.1
Reserved—Bits 15 – 8
This bit field is reserved or not implemented. It is read as 0 and cannot be modified by writing.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-31
Analog-to-Digital Converter (ADC)
28.5.6.2
Disable Sample (DS n )—Bits 7 – 0
The respective SAMPLE n field can be enabled or disabled where n = 0
–
7.
• 0 = Enable SAMPLE n
• 1 = Disable SAMPLE n and all subsequent samples. Which samples are actually disabled will depend on the conversion mode, sequential/parallel, and the value of SIMULT.
28.5.7
Status Register (STAT)
This register provides the current status of the ADC module. RDY n bits are cleared by reading their corresponding result (ADRSLT n ) registers. HLMTI and LLMTI bits are cleared by writing 1 to each asserted bit in the ADC limit status (ADLSTAT) register. Likewise, the ZCI bit, bit-10, is cleared by writing 1 to each asserted bit in the ADC zero crossing status (ADZCSTAT) register. The EOSI n bits are cleared by writing 1 to them. Please see Figure 23-25 for more information regarding the operation of interrupts.
Except for CIP0 and CIP1 all bits in the STAT Register are sticky—once set to a 1 state, they require some specific action to clear them. They are not cleared automatically on the next scan sequence.
IPSBAR
Offset:
0x19_0006 (STAT)
15 14 13
R CIP0 CIP1 0
W
Reset 0 0 0 0 0 0
Access: read/write
12 11
EOSI1 EOSI0
10 9 8 7 6 5 4 3 2 1 0
ZCI LLMTI HLMTI RDY7 RDY6 RDY5 RDY4 RDY3 RDY2 RDY1 RDY0
0 0 0 0 0 0 0 0 0
Figure 28-20. Status (STAT) Register
0
28.5.7.1
Conversion in Progress 0 (CIP0)—Bit 15
This bit indicates when a scan is in progress.
• 0 = Idle state
• 1 = A scan cycle is in progress. The ADC will ignore all sync pulses or start commands
This bit supports any sequential scan or parallel scan with SIMULT=1. When executing a parallel scan with SIMULT = 0, this bit services the scan of converter A while the CIP1 bit services the scan of converter
B.
28.5.7.2
Conversion in Progress 1 (CIP1)—Bit 14
This bit indicates when a scan is in progress.
• 0 = Idle state
• 1 = A scan cycle is in progress. The ADC will ignore all sync pulses or start commands
This refers only to a B converter scan in non-simultaneous (SIMULT=0) parallel scan modes.
28-32
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
28.5.7.3
Reserved—Bit 13
This bit is reserved or not implemented. It is read as 0 and cannot be modified by writing.
28.5.7.4
End of Scan Interrupt 1 (EOSI1)—Bit 12
This bit indicates whether a scan of analog inputs have been completed since the last read of the STAT register or a reset. The EOSI1 bit is cleared by writing 1 to it. This bit cannot be set by software.
• 0 = A scan cycle has not been completed, no end of scan IRQ pending
• 1 = A scan cycle has been completed, end of scan IRQ pending
In looping scan modes, this interrupt is triggered at the completion of each iteration of the loop. This interrupt is triggered only by the completion of a B converter scan in non-simultaneous (SIMULT=0) parallel scan modes. In this case the EOSI0 interrupt is triggered when converter A completes its scan.
28.5.7.5
End of Scan Interrupt 0 (EOSI0)—Bit 11
This bit indicates whether a scan of analog inputs has been completed since the last read of the STAT register or a reset. The EOSI0 bit is cleared by writing 1 to it. This bit cannot be set by software. EOSI0 is the preferred bit to poll for scan completion if interrupts are not enabled.
• 0 = A scan cycle has not been completed, no end of scan IRQ pending
• 1 =A scan cycle has been completed, end of scan IRQ pending
In looping scan modes, this interrupt is triggered at the completion of each iteration of a loop.
This interrupt is triggered upon the completion of any sequential scan or parallel scan with SIMULT=1.
When executing parallel scans with SIMULT = 0, this interrupt is triggered when converter A completes its scan while the EOSI1 interrupt services converter B.
28.5.7.6
Zero Crossing Interrupt (ZCI)—Bit 10
This bit is asserted at the completion of an individual conversion experiencing a zero crossing enabled in the ADC zero crossing control (ADZCC) register. The bit is set as soon as an enabled zero crossing event occurs rather than at the end of the ADC scan.
The ZCI bit is cleared by writing 1 to all active ZCS[7:0] bits in the ADZCSTAT register.
• 0 = No ZCI interrupt request
• 1 = Zero crossing encountered; IRQ pending if ZCIE is set
28.5.7.7
Low Limit Interrupt (LLMTI)—Bit 9
If any low limit (LOLIM n ) register is enabled by having a value other than $0000, low limit checking is enabled. This bit is set at the completion of an individual conversion which may or may not be the end of a scan.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-33
Analog-to-Digital Converter (ADC)
The LLMTI bit is cleared by writing 1 to all active LLS[7:0] bits in the ADLSTAT register.
• 0 = No low limit interrupt request
• 1 = Low limit exceeded, IRQ pending if LLMTIE is set
28.5.7.8
High Limit Interrupt (HLMTI)—Bit 8
If any high limit (HILIM n ) register is enabled by having a value other than 0x7FF8, high limit checking is enabled. This bit is set at the completion of an individual conversion which may or may not be the end of a scan.
The HLMTI bit is cleared by writing 1 to all active HLS[7:0] bits in the ADLSTAT register.
• 0 = No high limit interrupt request
• 1 = High limit exceeded, IRQ pending if HLMTIE is set
28.5.7.9
Ready Sample 7 – 0 (RDY n )—Bits 7 – 0
These bits indicate samples 7-0 are ready to be read. The RDY n bits are set as the individual channel conversions are completed and stored in a ADRSLT n register. These bits are cleared after a read from the corresponding ADC results (ADRSLT n ) register. If polling the RDY n bits to determine if a particular sample is executed, care should be taken to not start a new scan until all enabled samples are completed.
NOTE
RDY n bits can be cleared when the debugger reads the corresponding
Results register during a debug session.
• 0 = Sample not ready or has been read
• 1 = Sample ready to be read
illustrates how five interrupts sources are combined into three entries in the interrupt vector table.
EOSI0
EOSIE0
EOSI1
EOSIE1
ZCI
ZCIE
LLMTI
LLMTIE
HLMTI
HLTMIE
ADCA Conversion Complete
(ADC_CC0_INT)
ADCB Conversion Complete
(ADC_CC1_INT)
ADC Zero Crossing or Limit Error
(ADC_ERR_INT)
Figure 28-21. ADC Interrupt Sources
28-34
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
28.5.8
Limit Status Register (ADLSTAT)
The ADC limit status (ADLSTAT) register latches in the result of the comparison between the result of the sample in the ADRSLT n register and the respective limit register, HILIM n or LOLIM n .
For example, if the result for ADRSLT0 is greater than the value programmed into the HILIM0, then set the HLS0 bit to 1. An interrupt is generated if the HLMTIE bit is set in CTRL1. These bits are sticky—once set, the bits require a specific modification to clear them. They are not cleared automatically by subsequent conversions. A bit may only be cleared by writing a value of one to that specific bit.
IPSBAR
Offset:
0x19_0007 (ADLSTAT) Access: read/write
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
R
HLS7 HLS6 HLS5 HLS4 HLS3 HLS2 HLS1 HLS0 LLS7 LLS6 LLS5 LLS4 LLS3 LLS2 LLS1 LLS0
W
Reset 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Figure 28-22. Limit Status (ADLSTAT) Register
28.5.9
Zero Crossing Status Register (ADZCSTAT)
The ADC zero crossing status (ADZCSTAT) register latches in the result of the comparison between the current result of the sample and the previous result of the same results register. For example, if the result for the channel programmed in SAMPLE0 changes sign from the previous conversion, and the respective
ZCE bit in ADZCC register is set to 11b (any edge change), then set the ZCS0 bit to 1. An interrupt is generated if the ZCIE bit is set in the CTRL1 register. These bits are sticky—once set, they require a write to clear them. They are not cleared automatically by subsequent conversions. A bit may only be cleared by writing a value of 1 to that specific bit.
IPSBAR
Offset:
0x19_0008 (ADZCSTAT)
W
15
R 0
Reset 0
14
0
0
13
0
12
0
11
0
10
0
9
0
8
0
7 6 5 4
0 0 0 0 0 0 0 0 0 0
Figure 28-23. Zero Crossing Status (ADZCSTAT) Register
3
ZCS7 ZCS6 ZCS5 ZCS4 ZCS3 ZCS2 ZCS1 ZCS0
0
Access: read/write
2
0
1
0
0
0
28.5.9.1
Reserved—Bits 15 – 8
This bit field is reserved or not implemented. It is read as 0 and cannot be modified by writing.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-35
Analog-to-Digital Converter (ADC)
28.5.9.2
Zero Crossing Status (ZCS[7:0])—Bits 7 – 0
The zero crossing condition is determined by examining the ADC value after it has been adjusted by the offset for the result register. Please see
Each bit of the register is cleared by writing 1 to that register bit.
• 0 = a. A sign change did not occur in a comparing the current ADRSLT n value and the previous
ADRSLT n value, or b. Zero crossing control is disabled for sample n in the ADC zero crossing control
(ADZCC) register
• 1 = In a comparison between the current channelx result and the previous channelx result, a sign change condition occurred as defined in the ADC zero crossing control (ADZCC) register
28.5.10 Result 0-7 Registers (ADRSLT0–7)
The 8 result registers contain the converted results from a scan. The SAMPLE0 result is loaded into
ADRSLT0 register, SAMPLE1 result in ADRSLT1 register, and so on. In a simultaneous parallel scan mode, the first channel pair, designated by SAMPLE0 and SAMPLE4 in register LIST1/2, is stored in
ADRSLT0 and ADRSLT4, respectively.
When writing to this register, only the RSLT portion of the value written is used. This value is modified as shown in
and the result of the subtraction is stored. The SEXT bit is only set as a result of this subtraction and is not directly determined by the value written.
ADC Result Register 0
—
Address:
ADC Result Register 1
—
Address:
ADC Result Register 2
—
Address:
ADC Result Register 3
—
Address:
ADC Result Register 4
—
Address:
ADC Result Register 5
—
Address:
ADC Result Register 6
—
Address:
ADC Result Register 7
—
Address:
ADC_BASE + $12
ADC_BASE + $14
ADC_BASE + $16
ADC_BASE + $18
ADC_BASE + $1A
ADC_BASE + $1C
ADC_BASE + $1E
ADC_BASE + $20
IPSBAR
Offset:
0x19_0009 - 0x19_0010 (ADRSLT0–7)
14 13 12 11 10 15
R SEXT
W
Reset 0 0 0 0
9
RSLT
8 7 6 5
0 0 0 0 0 0 0
Figure 28-24. Result (ADRSLT0
–
7) Registers
0
4 3
0
Access: read/write
2
0
0
1
0
0
0
0
0
28-36
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
28.5.10.1 Sign Extend (SEXT)—Bit 15
SEXT is the sign-extend bit of the result. When the SEXT bit is set to 1, it implies a negative result. When the SEXT bit is set to 0, it implies a positive result. If only positive results are required, then the respective
ADC Offset (ADOFS n ) register must be set to a value of 0.
28.5.10.2 Digital Result of the Conversion (RSLT)—Bits 14 – 3
RSLT can be interpreted as either a signed integer or a signed fixed point (fractional) number. As a fixed point number, the RSLT can be used directly. As a signed integer, one has the option to right shift with sign extend (ASR) three places to fit it into the range [0,4095]. Or one can accept the number as presented in the register, knowing there are missing codes, because the lower three LSBs are always zero.
Negative results, SEXT = 1, are always presented in twos complement format. If it is a requirement of an application that the result registers always be positive, the offset register (ADOFS n) must always be set to
0.
The interpretation of the numbers programmed into the ADC limit and offset (LOLIM n , HILIM n , and
ADOFS n ) registers should match your interpretation of the result register.
28.5.10.3 Test Data (Test_Data)—Bits 14 – 3
When the ADC is stopped or in power-down mode, this field can be written by accessing the register in the memory map. Please see
Section 28.4.3 “ADC Data Processing ” more information.
28.5.10.4 Reserved—Bits 2 – 0
This bit field is reserved or not implemented. It is read as 0 and cannot be modified by writing.
28.5.11 Low and High Limit Registers (LOLIM0-7 and HILIM0-7)
Each ADC sample is compared against the values in the limit registers. The comparison is based upon the
raw conversion value before the offset correction is applied. Refer to Figure 28-7 .
ADC limit registers
(LOLIM n and HILIM n) correspond to results (ADRSLT n ) registers. The high limit register is used for the comparison of result > high limit. The low limit register is used for the comparison of result < low limit.
The limit checking can be disabled by programming the respective limit register with 0x7FF8 for the high limit and 0x0000 for the low limit. At reset, limit checking is disabled.
ADC Low Limit Register 0
—
Address: ADC_BASE + $22
ADC Low Limit Register 1
—
Address: ADC_BASE + $24
ADC Low Limit Register 2
—
Address: ADC_BASE + $26
ADC Low Limit Register 3
—
Address: ADC_BASE + $28
ADC Low Limit Register 4
—
Address: ADC_BASE + $2A
ADC Low Limit Register 5
—
Address: ADC_BASE + $2C
ADC Low Limit Register 6
—
Address: ADC_BASE + $2E
ADC Low Limit Register 7
—
Address: ADC_BASE + $30
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-37
Analog-to-Digital Converter (ADC)
IPSBAR
Offset:
0x19_0011 - 0x19_0018 (LOLIM0–7)
14 13 12 11 15
R 0
W
Reset 0 0 0 0 0
10
0
9
LLMT
8 7 6
0 0
Figure 28-25.
0 0
ADC High Limit Register 0
—
Address: ADC_BASE + $32
ADC High Limit Register 1
—
Address: ADC_BASE + $34
ADC High Limit Register 2
—
Address: ADC_BASE + $36
ADC High Limit Register 3
—
Address: ADC_BASE + $38
ADC High Limit Register 4
—
Address: ADC_BASE + $3A
ADC High Limit Register 5
—
Address: ADC_BASE + $3C
ADC High Limit Register 6
—
Address: ADC_BASE + $3E
ADC High Limit Register 7
—
Address: ADC_BASE + $40
5
0
IPSBAR
Offset:
0x19_0019 - 0x19_0020 (HILIM0–7)
14 13 12 11 15
R 0
W
Reset 0 0 0 0
10 9
HLMT
8 7 6 5
0 0 0 0 0 0 0
Figure 28-26. High Limit Register (HILIM0–7)
4
0
4
0
3
0
3
0
Access: read/write
2
0
0
2
0
1
0
1
0
0
0
0
0
Access: read/write
0
0
0
0
0
28.5.12 Offset Registers (ADOFS0–7)
Value of the offset (ADOFS n ) register is used to correct the ADC result before it is stored in the ADRSLT n registers.
ADC Offset Register 0
—
Address:
ADC Offset Register 1
—
Address:
ADC Offset Register 2
—
Address:
ADC Offset Register 3
—
Address:
ADC Offset Register 4
—
Address:
ADC Offset Register 5
—
Address:
ADC Offset Register 6
—
Address:
ADC Offset Register 7
—
Address:
ADC_BASE + $42
ADC_BASE + $44
ADC_BASE + $46
ADC_BASE + $48
ADC_BASE + $4A
ADC_BASE + $4C
ADC_BASE + $4E
ADC_BASE + $50
28-38
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
IPSBAR
Offset:
0x19_0021 - 0x19_0028 (ADOFS0–7)
14 13 12 11 15
R 0
W
Reset 0 0 0 0 0
10
0
9
OFFSET
0
8
0
7
0
6
0
5
0
Figure 28-27. Offset 0-7 (ADOFS0-7) Registers
4
0
3
0
Access: read/write
2
0
0
1
0
0
0
0
0
The offset value is subtracted from the ADC result. In order to obtain unsigned results, the respective offset register should be programmed with a value of $0000, thus giving a result range of $0000 to $7FF8.
28.5.13 Power Control Register (POWER)
This register controls the power management features of the ADC module. There are manual power-down control bits for the two ADC converters and the shared voltage reference generator. There are also 5 distinct power modes with related controls:
1. Powered-down state
Each converter and the voltage reference generator can individually be put into a powered down state. When powered down, the unit consumes no power. Results of scans referencing a powered down converter are undefined. The voltage reference generator and at least one converter must be powered up to use the ADC module.
2. Manual power-down controls
Each converter and the voltage reference generator have a manual power control bit capable of forcing that component into the power down state. Also, each converter and the voltage reference generator can be powered up/down automatically as part of ADC operation.
3. Idle state
The ADC module is idle when neither of the two converters has a scan in process.
4. Active state
The ADC module is active when at least one of the two converters has a scan in process.
5. Current mode
• Normal current mode is used to power the converters at clock rates above 100 kHz.
• Standby current mode uses less power and is engaged only when the ADC clock is at 100 kHz. The current mode active does not affect the number of ADC clock cycles required to do a conversion or the accuracy of a conversion. The ADC module may change the current mode when idle as part of the power saving strategy. Both converters will be in the same current mode at all times.
In addition to the power modes, there is startup delay:
• Auto power-down and auto standby power modes cause a startup delay when the ADC module goes between the idle and active states to allow time to switch clocks or power configurations. The number of ADC clocks used in the startup delay is defined by the PUDELAY field.
See the discussion of power modes in the Functional Description
Section 28.4 “Functional Description ”
detailed description of the clocking system and the control of current mode.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-39
Analog-to-Digital Converter (ADC)
IPSBAR
Offset:
0x19_0029 (POWER)
15
R
W
ASB
Reset 0
14
0
0
13 12 11 10
0 PSTS2 PSTS1 PSTS0
0 0
9 8 7 6
PUDELAY
5
0 0 0 0 1 1 0
Figure 28-28. Power Control (POWER) Register
4
1
Access: read/write
3 2 1 0
APD PD2 PD1 PD0
0 1 1 1
28.5.13.1 Auto Standby (ASB)—Bit 15
The ASB bit selects auto standby mode. ASB is ignored if APD is 1. When the ADC is idle, auto standby mode selects the standby clock as the ADC clock source and puts the converters into standby current mode.
At the start of any scan, the conversion clock is selected as the ADC clock and a delay of PUDELAY ADC clock cycles is imposed for current levels to stabilize. After this delay, the ADC will initiate the scan.
When the ADC returns to the idle state, the standby clock is again selected and the converters revert to the standby current state.
• 0 = Auto standby mode disabled
• 1 = Auto standby mode enabled
28.5.13.2 Reserved—Bits 14–13
This bit field is reserved or not implemented. It is read as 0 and cannot be modified by writing.
28.5.13.3 Voltage Reference Power Status 2 (PSTS2)—Bit 12
PSTS2 is a read-only bit. It simply reflects whether the voltage reference circuit is currently enabled.
• 0 = Voltage reference circuit is currently powered up
• 1 = Voltage reference circuit is currently powered down
28.5.13.4 Converter B Power Status 1 (PSTS1)—Bit 11
PSTS1 is a read-only bit. It is asserted immediately following a write of 1 to PD1. It is deasserted
PUDELAY ADC clock cycles after writing 0 to PD1 if APD is 0. This bit can be read as a status bit to determine when the ADC is ready for operation. During auto power-down mode, this bit indicates the current powered state of converter B.
• 0 = ADC converter B is currently powered up
• 1 = ADC converter B is currently powered down
28.5.13.5 Converter A Power Status 0 (PSTS0)—Bit 10
PSTS0 is a read-only bit. It is asserted immediately following a write of 1 to PD0. It is deasserted
PUDELAY ADC clock cycles after writing 0 to PD0 if APD is 0. This bit can be read as a status bit to determine when the ADC is ready for operation. During auto power-down mode, this bit indicates the current powered state of converter A.
28-40
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
• 0 = ADC converter A is currently powered up
• 1 = ADC converter A is currently powered down
28.5.13.6 Power-Up Delay (PUDELAY)—Bits 9–4
This 6-bit field determines the number of ADC clocks provided to power-up an ADC converter (after setting PD0 or PD1 to 0) before allowing a scan to start. It also determines the number of ADC clocks of delay provided in auto power-down (APD) and auto standby (ASB) modes between when the ADC goes from the idle to active state and when the scan is allowed to start. The default value is 13 ADC clocks.
Accuracy of the initial conversions in a scan will be degraded if PUDELAY is set to too small a value.
NOTE
PUDELAY defaults to a value typically sufficient for any power mode. The latency of a scan can be reduced by reducing PUDELAY to the lowest value for which accuracy is not degraded. Please refer to the Device Data Sheet for further details.
28.5.13.7 Auto Power-Down (APD)—Bit 3
Auto power-down mode powers down converters when not in use for a scan. APD takes precedence over
ASB. When a scan is started in APD mode, a delay of PUDELAY ADC clock cycles is imposed during which the needed converter(s), if idle, are powered up. The ADC will then initiate a scan equivalent to when APD is not active. When the scan is completed, the converter(s) are powered down again.
• 0 = Auto power-down mode is not active
• 1 = Auto power-down mode is active
NOTE
If ASB or APD is asserted while a scan is in progress, that scan is unaffected and the ADC will wait to enter its low power state until after all conversions are complete and both ADCs are idle.
NOTE
ASB and APD are not useful in looping modes. The continuous nature of scanning means the low power state can never be entered.
28.5.13.8 Power-Down Control for Voltage Reference Circuit 2 (PD2)—Bit 2
This bit controls the power-down of the ADC’s voltage reference current.
• 0 = Manually power-up voltage reference circuit
• 1 = Power-down voltage reference circuit is controlled by PD0 and PD1 (default)
The voltage reference circuit is shared by both converters. When PD2=1 the voltage reference will be activated whenever PD1 or PD0 are powered up. It is not usually necessary to modify this bit, since powering down both converter A and converter B will automatically power-down the voltage reference.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-41
Analog-to-Digital Converter (ADC)
28.5.13.9 Manual Power-Down for Converter B (PD1)—Bit 1
This bit forces ADC converter B to power-down.
• 0 = Power-up ADC converter B
• 1 = Power-down ADC converter B
Asserting PD1 powers down converter B immediately. The results of a scan using converter B will be invalid while PD1 is asserted. When PD1 is cleared, converter B is either continuously powered up (APD
= 0) or automatically powered up when needed (APD=1).
When clearing PD1 in any power mode except auto power-down (APD=1), wait PUDELAY ADC clock cycles before initiating a scan to stabilize power levels within the converter. The PSTS1 bit can be polled to determine when the PUDELAY time has elapsed. Failure to follow this procedure can result in loss of accuracy of the first two samples.
28.5.13.10 Manual Power-Down for Converter A (PD0)—Bit 0
This bit forces ADC converter A to power-down.
• 0 = Power-up ADC converter A
• 1 = Power-down ADC converter A
Asserting PD0 powers down converter A immediately. The results of a scan using converter A will be invalid while PD0 is asserted. When PD0 is cleared, converter A is either continuously powered up (APD
= 0) or automatically powered up when needed (APD=1).
When clearing PD0 in any power mode except auto power-down (APD=1), wait PUDELAY ADC clock cycles before initiating a scan to stabilize power levels within the converter. The PSTS0 bit can be polled to determine when the PUDELAY time has elapsed.
NOTE
Failure to follow this procedure can result in loss of accuracy of the first two samples.
28.5.14 Voltage Reference Register (CAL)
In earlier series, this register supported ADC calibration and had a different name. Improvements in ADC performance have eliminated the need for on-chip calibration support, hence the new name.
Access: read/write IPSBAR
Offset:
0x19_002A (CAL)
15 14
R
SEL_VREFH SEL_VREFL
W
Reset 0 0
13
0
12
0
11
0
10
0
9
0
8
0
7
0
6
0
5
0
0 0 0 0 0 0 0 0 0
Figure 28-29. Voltage Reference (CAL) Register
4
0
0
3
0
0
2
0
0
1
0
0
0
0
0
28-42
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Analog-to-Digital Converter (ADC)
28.5.14.1 Select V
REFH
Source (SEL_VREFH)—Bit 15
This bit selects the source of the V
REFH
reference for conversions.
• 0 = Internal VR
X
• 1 = AN2
28.5.14.2 Select V
REFL
Source (SEL_VREFL)—Bit 14
This bit selects the source of the V
REFL
reference for conversions.
• 0 = Internal VR
X
• 1 = AN6
28.5.14.3 Reserved—Bits 13–0
This bit field is reserved or not implemented. It is read as 0 and cannot be modified by writing.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
28-43
Analog-to-Digital Converter (ADC)
28-44
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 29
Pulse Width Modulation (PWM) Module
29.1
Introduction
This chapter describes the configuration and operation of the pulse width modulation (PWM) module. It includes a block diagram, programming model, and functional description.
29.1.1
Overview
The PWM module shown in Figure 29-1 , generates a synchronous series of pulses having programmable
period and duty cycle. With a suitable low-pass filter, the PWM can be used as a digital-to-analog converter.
Internal Bus
Clock (f sys/2
)
Clock select
PWM Clocks
Control
Enable
Polarity
Alignment
PWM Channels
Channel 7
Period and Duty Counter
Channel 6
Period and Duty Counter
Channel 5
Period and Duty Counter
Channel 4
Period and Duty Counter
Channel 3
Period and Duty Counter
Channel 2
Period and Duty Counter
Channel 1
Period and Duty Counter
Channel 0
Period and Duty Counter
PWMOUT7
PWMOUT5
PWMOUT3
PWMOUT1
Figure 29-1. PWM Block Diagram
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
29-1
Pulse Width Modulation (PWM) Module
Main features include the following:
• Double-buffered period and duty cycle
• Left- or center-aligned outputs
• Eight independent PWM modules. Notice that only the four odd PWM channel outputs are available on the device. The even channels can be used for concatenation purposes to generate
16-bit PWM for the odd channels.
• Byte-wide registers provide programmable duty cycle and period control
• Four programmable clock sources
NOTE
The GPIO module must be configured to enable the peripheral function of the appropriate pins (refer to Chapter 11, “General Purpose I/O Module ”) prior to configuring the PWM module.
29.2
Memory Map/Register Definition
This section describes the registers and control bits in the PWM module. There are eight independent
PWM modules, each with its own control and counter registers, although only four channels have an output signal. The memory map for the PWM is shown below.
Table 29-1. PWM Memory Map
IPSBAR Offset
1,2
Register
Width
(bits)
Access Reset Value Section/Page
0x001B_0000 PWM Enable Register (PWME)
0x001B_0001 PWM Polarity Register (PWMPOL)
0x001B_0002 PWM Clock Select Register (PWMCLK)
0x001B_0003 PWM Prescale Clock Select Register (PWMPRCLK)
8
8
8
8
R/W
R/W
R/W
R/W
0x00
0x00
0x00
0x00
0x001B_0004
0x001B_0005
0x001B_0008
0x001B_0009
PWM Center Align Enable Register (PWMCAE)
PWM Control Register (PWMCTL)
PWM Scale A Register (PWMSCLA)
PWM Scale B Register (PWMSCLB)
8
8
8
8
8
R/W
R/W
R/W
R/W
R/W
0x00
0x00
0x00
0x00
0x00
0x001B_000C + n n = 0–7
PWM Channel n Counter Register (PWMCNT n )
0x001B_0014 + n n = 0–7
PWM Channel n Period Register (PWMPER n )
0x001B_001C + n n = 0–7
PWM Channel n Duty Register (PWMDTY n )
8
8
R/W
R/W
0xFF
0xFF
0x001B_0024 PWM Shutdown Register (PWMSDN) 8 R/W 0x00
1
Addresses not assigned to a register and undefined register bits are reserved for expansion. Write accesses to these reserved address spaces and reserved register bits have no effect.
29-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Pulse Width Modulation (PWM) Module
2
32-bit access to any of these registers will result in a bus transfer error (see Section 11.2.7, “SCM Interrupt Status Register
(SCMISR)” ).
29.2.1
PWM Enable Register (PWME)
Each PWM channel has an enable bit (PWME n ) to start its waveform output. While in run mode, if all four
PWM output channels are disabled (PWME[7:0] = 0), the prescaler counter shuts off for power savings.
See Section 29.3.2.1, “PWM Enable”
for more information.
Address: 0x001B_0000 (PWME)
7 6
0 R
W
Reset:
PWME7
0
5
PWME5
4
0
3
PWME3
0 0 0 0
Figure 29-2. PWM Enable Register (PWME)
0
2
0
Access: User Read/Write
1 0
0
PWME1
0 0
Table 29-2. PWME Field Descriptions
Field Description
7
PWME5
PWM channel 7 output enable. If enabled, the PWM signal becomes available at PWMOUT7 when its corresponding clock source begins its next cycle.
0 PWM output disabled
1 PWM output enabled
6 Reserved, should be cleared.
5
PWME5
PWM channel 5 output enable. If enabled, the PWM signal becomes available at PWMOUT5 when its corresponding clock source begins its next cycle.
0 PWM output disabled
1 PWM output enabled
4 Reserved, should be cleared.
3
PWME3
PWM channel 3 output enable. If enabled, the PWM signal becomes available at PWMOUT3 when its corresponding clock source begins its next cycle.
0 PWM output disabled
1 PWM output enabled
2 Reserved, should be cleared.
1
PWME1
PWM channel 1 output enable. If enabled, the PWM signal becomes available at PWMOUT1 when its corresponding clock source begins its next cycle.
0 PWM output disabled
1 PWM output enabled
0 Reserved, should be cleared.
29.2.2
PWM Polarity Register (PWMPOL)
The starting polarity of each PWM channel waveform is determined by the associated PWMPOL[PPOL n ] bit. If the polarity is changed while a PWM signal is being generated, a truncated or stretched pulse can occur during the transition.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
29-3
Pulse Width Modulation (PWM) Module
Address: 0x001B_0001 (PWMPOL)
7 6
0 R
W
Reset:
PPOL7
0
5
PPOL5
4
0
3
PPOL3
0 0 0 0
Figure 29-3. PWM Polarity Register (PWMPOL)
0
2
0
Access: User Read/Write
1 0
0
PPOL1
0 0
Table 29-3. PWMPOL Field Descriptions
Field Description
7,5,3,1
PPOL n
PWM channel n polarity.
0 PWM channel n output is low at the beginning of the period, then goes high when the duty count is reached
1 PWM channel n output is high at the beginning of the period, then goes low when the duty count is reached
6,4,2,0 Reserved, should be cleared.
29.2.3
PWM Clock Select Register (PWMCLK)
Each PWM channel has the capability of selecting one of two clocks. For channels 1 and 5, the clock choices are clock A or SA. For channels 3 and 7, the choices are clock B or SB. The clock selection is done with the below PWMCLK[PCLK n ] control bits. If a clock select is changed while a PWM signal is being generated, a truncated or stretched pulse can occur during the transition.
Address: 0x001B_0002 (PWMCLK)
7 6
0 R
W
Reset:
PCLK7
0
5
PCLK5
4
0
3
PCLK3
2
0
0 0 0 0 0
Figure 29-4. PWM Clock Select Register (PWMCLK)
Access: User Read/Write
1 0
0
PCLK1
0 0
29-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Pulse Width Modulation (PWM) Module
Table 29-4. PWMCLK Field Descriptions
Field Description
7,5,3,1
PCLK n
PWM channel n clock select. Selects between one of two clock sources for each PWM channel. See
“PWM Prescale Clock Select Register (PWMPRCLK)”
and Section 29.2.7, “PWM Scale A Register (PWMSCLA)”
for more information on how the different clock rates are generated. The even-numbered channels’ clock select has no effect when the corresponding PWMCTL[CON n(n+1) ] bit is set. For example, if PWMCTL[CON01] = 1 then
PWMCLK[PCLK0] has no affect.
0
1
PCLK7
(PCLK7 Clock
Source)
B
SB
PCLK5
(PWM5 Clock
Source)
A
SA
PCLK3
(PWM3 Clock
Source)
B
SB
PCLK1
(PWM1 Clock
Source)
A
SA
6,4,2,0 Reserved, should be cleared.
29.2.4
PWM Prescale Clock Select Register (PWMPRCLK)
The PWMPRCLK register selects the prescale clock source for clocks A and B independently. If the clock prescale is changed while a PWM signal is being generated, a truncated or stretched pulse can occur during the transition.
Address: 0x001B_0003 (PWMPRCLK)
6 7
0 R
W
Reset:
5
PCKB
4 3
0
2
Access: User Read/Write
1
PCKA
0 0 0 0 0 0
Figure 29-5. PWM Prescale Clock Select Register (PWMPRCLK)
0
0
0
Table 29-5. PWMPRCLK Field Descriptions
Field
7
6–4
PCKB
Description
Reserved, should be cleared.
Clock B prescalar select. These three bits control the rate of Clock B which can be used for PWM channels 3 and 7.
000
001
...
111
Internal bus clock
÷
2
0
Internal bus clock
÷
2
1
...
Internal bus clock
÷
2
7
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
29-5
Pulse Width Modulation (PWM) Module
Table 29-5. PWMPRCLK Field Descriptions (continued)
Field
3
2–0
PCKA
Description
Reserved, should be cleared.
Clock A prescalar select. These three bits control the rate of Clock A which can be used for PWM channels 1 and 5.
PCKA
000
001
...
111
Clock A Rate
Internal bus clock
÷
2
0
Internal bus clock
÷
2
1
...
Internal bus clock
÷
2
7
29.2.5
PWM Center Align Enable Register (PWMCAE)
The PWMCAE register contains four control bits for the selection of center-aligned outputs or left-aligned outputs for each PWM channel. Write these bits only when the corresponding channel is disabled. See
Section 29.3.2.5, “Left-Aligned Outputs”
and Section 29.3.2.6, “Center-Aligned Outputs”
for a more detailed description of the PWM output modes.
Address: 0x001B_0004 (PWMCAE)
7 6
0 R
W
Reset:
CAE7
0
5
CAE5
4
0
3
CAE3
2
0
0 0 0 0 0
Figure 29-6. PWM Center Align Enable Register (PWMCAE)
Access: User Read/Write
1
CAE1
0
0
0
0
Table 29-6. PWMCAE Field Descriptions
Field Description
7,5,3,1
CAE n
Center align enable for channel n . The even-numbered channels’ center align enable has no effect when the corresponding PWMCTL[CON n(n+1) ] bit is set. For example, if PWMCTL[CON01] = 1 then PWMCAE[CAE0] has no affect.
0 Channel n operates in left-aligned output mode
1 Channel n operates in center-aligned output mode
6,4,2,0 Reserved, should be cleared.
29.2.6
PWM Control Register (PWMCTL)
The PWMCTL register provides various control of the PWM module. Change the CON n(n+1) bits only when both corresponding channels are disabled. See
Section 29.3.2.7, “PWM 16-Bit Functions”
for a more detailed description of the concatenation function.
29-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Pulse Width Modulation (PWM) Module
Address: 0x001B_0005 (PWMCTL)
7 6
R
W
Reset:
CON67
0
CON45
5
CON23
4
CON01
3
PSWAI
2
PFRZ
0 0 0 0
Figure 29-7. PWM Control Register (PWMCTL)
0
Access: User Read/Write
1
0
0
0
0 0
Table 29-7. PWMCTL Field Descriptions
Field Description
7
CON67
Concatenates PWM channels 6 and 7 to form one 16-bit PWM channel.
0 Channels 6 and 7 are separate 8-bit PWMs. Note that there is no PWM 6 output.
1 Concatenate PWM 6 and 7. Channel 6 becomes the high order byte and channel 6 the low order byte. PWMOUT7 is the output for this 16-bit PWM signal, and PWMOUT6 is disabled. The channel 7 clock select, polarity, center align enable, and enable bits control this concatenated output.
6
CON45
Concatenates PWM channels 4 and 5 to form one 16-bit PWM channel.
0 Channels 4 and 5 are separate 8-bit PWMs. Note that there is no PWM 4 output.
1 Concatenate PWM 4 and 5. Channel 4 becomes the high order byte and channel 5 the low order byte. PWMOUT5 is the output for this 16-bit PWM signal, and PWMOUT4 is disabled. The channel 5 clock select, polarity, center align enable, and enable bits control this concatenated output.
5
CON23
Concatenates PWM channels 2 and 3 to form one 16-bit PWM channel.
0 Channels 2 and 3 are separate 8-bit PWMs. Note that there is no PWM 2 output.
1 Concatenate PWM 2 and 3. Channel 2 becomes the high order byte and channel 3 the low order byte. PWMOUT3 is the output for this 16-bit PWM signal, and PWMOUT2 is disabled. The channel 3 clock select, polarity, center align enable, and enable bits control this concatenated output.
4
CON01
Concatenates PWM channels 0 and 1 to form one 16-bit PWM channel.
0 Channels 0 and 1 are separate 8-bit PWMs. Note that there is no PWM 0 output.
1 Concatenate PWM 0 and 1. Channel 0 becomes the high order byte and channel 1 the low order byte. PWMOUT1 is the output for this 16-bit PWM signal, and PWMOUT0 is disabled. The channel 1 clock select, polarity, center align enable, and enable bits control this concatenated output.
3
PSWAI
PWM stops in doze mode. Disables the input clock to the prescaler while in doze mode.
0 Allow the clock to the prescaler while in doze mode
1 Stop the input clock to the prescaler whenever the core is in doze mode
2
PFRZ
1–0
PWM counters stop in debug mode (BKPT asserted).
0 Allow PWM counters to continue while in debug mode
1 Disable PWM input clock to the prescaler when the core is in debug mode. Useful for emulation as it allows the
PWM function to be suspended.
Reserved, should be cleared.
29.2.7
PWM Scale A Register (PWMSCLA)
PWMSCLA is the programmable scale value used in scaling clock A to generate clock SA. Clock SA is generated with the following equation:
Clock SA = Eqn. 29-1
Any value written to this register will cause the scale counter to load the new scale value (PWMSCLA).
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
29-7
Pulse Width Modulation (PWM) Module
Address: 0x001B_0008 (PWMSCLA)
7 6
R
W
Reset: 0
5 4
SCALEA
3 2
0 0 0 0 0
Figure 29-8. PWM Scale A Register (PWMSCLA)
Table 29-8. PWMSCLA Field Descriptions
Field Description
7–0
SCALEA
Part of divisor used to form Clock SA from Clock A.
SCALEA Value
0x00
0x01
0x02
...
0xFF
256
1
2
...
255
Access: User Read/Write
1 0
0 0
29.2.8
PWM Scale B Register (PWMSCLB)
PWMSCLB is the programmable scale value used in scaling clock B to generate clock SB. Clock SB is generated according to the following equation:
Clock SB = Eqn. 29-2
Any value written to this register will cause the scale counter to load the new scale value (PWMSCLB).
Address: 0x001B_0009 (PWMSCLB)
7 6
R
W
Reset: 0
5 4
SCALEB
3 2
0 0 0 0 0
Figure 29-9. PWM Scale B Register (PWMSCLB)
Access: User Read/Write
1 0
0 0
29-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Table 29-9. PWMSCLB Field Descriptions
Field Description
7–0
SCALEB
Divisor used to form Clock SB from Clock B.
SCALEB Value
0x00
0x01
0x02
...
0xFF
256
1
2
...
255
Pulse Width Modulation (PWM) Module
29.2.9
PWM Channel Counter Registers (PWMCNT
n
)
Each channel has a dedicated 8-bit up/down counter that runs at the rate of the selected clock source,
PWMCLK[PCLK n ]. The user can read the counters at any time without affecting the count or the operation of the PWM channel. In left-aligned output mode, the counter counts from 0 to the value in the period register minus 1. In center-aligned output mode, the counter counts from 0 up to the value in the period register and then back down to 0. Therefore, given the same value in the period register, center-aligned mode is twice the period of left-aligned mode.
Any value written to the counter causes the counter to reset to 0x00, the counter direction to be set to up for center-aligned mode, the immediate load of both duty and period registers with values from the buffers, and the output to change according to the polarity bit.
The counter is also cleared at the end of the effective period (see
Section 29.3.2.5, “Left-Aligned Outputs”
and Section 29.3.2.6, “Center-Aligned Outputs”
for more details). When the channel is disabled
(PWME n =0), the PWMCNT n register does not count. When a channel is enabled (PWME n =1), the associated PWM counter starts at the count in the PWMCNT n register. For more detailed information on the operation of the counters, refer to
Section 29.3.2.4, “PWM Timer Counters.”
Address: 0x001B_000C (PWMCNT0)
0x001B_000D (PWMCNT1)
0x001B_000E (PWMCNT2)
0x001B_000F (PWMCNT3)
0x001B_0010 (PWMCNT4)
0x001B_0011 (PWMCNT5)
0x001B_0012 (PWMCNT6)
0x001B_0013 (PWMCNT7)
7 6
R
W
Reset: 0
5 4
COUNT
3 2
0 0 0 0 0
Figure 29-10. PWM Counter Registers (PWMCNT n )
Access: User Read/Write
1
0
0
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
29-9
Pulse Width Modulation (PWM) Module
Table 29-10. PWMCNT n Field Descriptions
Field Description
7–0
COUNT
Current value of the PWM up counter. Resets to zero when written.
29.2.10 PWM Channel Period Registers (PWMPER
n
)
for more information.
Calculating the output period depends on the output mode (center-aligned has twice the period as left-aligned mode) as well as PWMPER n . See the below equation:
PWM n period = Channel clock period × ( PWMCAE CAE n ] + 1 ) × PWMPER n Eqn. 29-3
For boundary case programming values (e.g. PWMPER n = 0x00), please refer to
.”
Address: 0x001B_0014 (PWMPER0)
0x001B_0015 (PWMPER1)
0x001B_0016 (PWMPER2)
0x001B_0017 (PWMPER3)
0x001B_0018 (PWMPER4)
0x001B_0019 (PWMPER5)
0x001B_001A (PWMPER6)
0x001B_001B (PWMPER7)
7 6
R
W
Reset: 1
5 4
PERIOD
3 2
1 1 1 1 1
Figure 29-11. PWM Period Registers (PWMPER n )
Table 29-11. PWMPER n Field Descriptions
Access: User Read/Write
1
1
0
1
Field Description
7–0
PERIOD
Period counter for the output PWM signal.
If PERIOD = 0x00, the PWM n output is always high (PPOL n =1) or always low (PPOL n =0). See
for other special cases.
29.2.11 PWM Channel Duty Registers (PWMDTY
n
)
The PWM duty registers determine the duty cycle of the associated PWM channel. To calculate the output duty cycle (high time as a percentage of period) for a particular channel:
Duty Cycle = – n ] –
PWMDTY n
PWMPER n
⎞
× 100% Eqn. 29-4
29-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Pulse Width Modulation (PWM) Module
For boundary case programming values (e.g. PWMDTY n = 0x00 or PWMDTY n > PWMPER n ), refer to
Section Section 29.3.2.8, “PWM Boundary Cases.”
Address: 0x001B_001C (PWMDTY0)
0x001B_001D (PWMDTY1)
0x001B_001E (PWMDTY2)
0x001B_001F (PWMDTY3)
0x001B_0020 (PWMDTY4)
0x001B_00241 (PWMDTY5)
0x001B_0022 (PWMDTY6)
0x001B_0023 (PWMDTY7)
7 6
R
W
Reset: 1
5 4
DUTY
3 2
1 1 1 1 1
Figure 29-12. PWM Duty Registers (PWMDTY n )
Access: User Read/Write
1
1
0
1
Table 29-12. PWMDTY n Field Descriptions
Field
7–0
DUTY
Description
Contains the duty value used to determine when a transition will occur on the PWM output signal. When a match occurs with the corresponding PWMCNT n register, the PWM output will toggle.
If DUTY = 0x00, the PWM n output is always low (PPOL n =1) or always high (PPOL n =0). See
for other special cases.
29.2.12 PWM Shutdown Register (PWMSDN)
The PWM shutdown register provides emergency shutdown functionality of the PWM module. The
PWMSDN[7:1] bits are ignored if PWMSDN[SDNEN] is cleared.
Address: 0x24 (PWMSDN)
R
W
Reset:
7
IF w1c
0
6
IE
5
0
RESTART
0
4
LVL
3
0
2
PWM7IN
0 0 0 0
Figure 29-13. PWM Shutdown Register (PWMSDN)
Access: Read/Write
1 0
PWM7IL
0
SDNEN
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
29-11
Pulse Width Modulation (PWM) Module
Table 29-13. PWMSDN Field Descriptions
Field Description
7
IF
6
IE
PWM interrupt flag. Any change in state of PWM7IN will be flagged by setting this bit. The flag is cleared by writing a ‘1’ to it. Writing ‘0’ has no effect.
0 No change in PWM7IN input
1 Change in PWM7IN input
PWM interrupt enable. An interrupt will be triggered to the device’s interrupt controller when PWMSDN[IF] is set.
0 Interrupt is disabled
1 Interrupt is enabled
5
RESTART
PWM restart. After setting the RESTART bit, the PWM channels start running after the corresponding counter resets to zero. Also, if emergency shutdown is cleared (after being set), the PWM outputs will restart after the corresponding counter resets to zero. This bit is self-clearing, so is always read as zero.
4
LVL
3
PWM shutdown output level. Describes the behavior of the PWM outputs when PWM7IN input is asserted and
PWMSDN[SDNEN] is set.
0 PWM outputs are forced to logic 0
1 PWM outputs are forced to logic 1
Reserved, should be cleared.
2
PWM7IN
PWM channel 7 input status. Reflects the current status of the PWMOUT7 pin. Read only.
1
PWM7IL
PWM channel 7 input polarity. If PWMSDN[SDNEN] is set, this bit sets the active level of the PWM 7 channel
0 PWM 7 input is active low
1 PWN 7 input is active high
0
SDNEN
PWM emergency shutdown enable. If set, the pin associated with PWM channel 7 is forced to input and the emergency shutdown feature is enabled.
0 Emergency shutdown is disabled
1 Emergency shutdown is enabled
29.3
Functional Description
29.3.1
PWM Clock Select
There are four available clocks—clock A, B, SA (scaled A), and SB (scaled B)—all based on the internal bus clock.
Clock A and B can be programmed to run at 1, 1/2,..., 1/128 times the internal bus clock. Clock SA and
SB use clock A and B respectively as an input and divides it further with a reloadable counter. The rates available for clock SA and SB are programmable to run at clock A and B divided by 2, 4,..., or 512. Each
PWM channel has the capability of selecting one of two clocks, either the prescaled clock (clock A or B)
and how the scaled clocks are created.
29-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Pulse Width Modulation (PWM) Module
Internal Bus
Clock (f sys/
)
PWMPRCLK
[PCKA]
PWMPRCLK
[PCKB]
PWMSCLB
Clock B
1
0
1
0
PCLR0
Clock to
PWM0
Clock to
PWM1
PWMSCLA
Clock SA
÷
2
PCLR1
PCLR4
1
0
Clock to
PWM4
1
0
Clock to
PWM5
Clock SB
PCLR5
PCLR2
1
÷
2 0
1
0
Clock to
PWM2
1
0
1
0
PCLR3
PCLR6
Clock to
PWM3
Clock to
PWM6
Clock to
PWM7
Figure 29-14. PWM Clock Select Block Diagram
PCLR7
29.3.1.1
Prescaled Clock (A or B)
The internal bus clock is the input clock to the PWM prescaler that can be disabled when the device is in debug mode by setting the PWMCTL[PFRZ] bit. This is useful for reducing power consumption and for emulation in order to freeze the PWM. The input clock is also disabled when all PWM channels are disabled (PWME n =0).
Clock A and B are scaled values of the input clock. The value is software selectable for both clock A and
B and has options of 1, 1/2,..., or 1/128 times the internal bus clock. The value selected for clock A and B is determined by the PWMPRCLK[PCKA n ] and PWMPRCLK[PCKB n ] bits.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
29-13
Pulse Width Modulation (PWM) Module
29.3.1.2
Scaled Clock (SA or SB)
The scaled A (SA) and scaled B (SB) clocks use clock A and B respectively as inputs, divide it further with a user programmable value, then divide this by 2. The rates available for clock SA are programmable to run at clock A divided by 2, 4,..., or 512. Similar rates are available for clock SB.
Clock SA equals clock A divided by two times the value in the PWMSCLA register:
Clock SA =
Similarly, clock SB is generated according to the following equation:
Eqn. 29-5
Clock SB = Eqn. 29-6
As an example, consider the case in which the user writes 0xFF into the PWMSCLA register. Clock A for this case is selected to be internal bus clock divided by 4. A pulse will occur at a rate of once every 255 × 4 bus cycles. Passing this through the divide by two circuit produces a clock signal of the internal bus clock divided by 2040. Similarly, a value of 0x01 in the PWMSCLA register when clock A is internal bus clock divided by 4 will produce an internal bus clock divided by 8 rate.
Writing to PWMSCLA or PWMSCLB causes the associated 8-bit down counter to be re-loaded.
Otherwise, when changing rates, the counter would have to count down to 0x01 before counting at the proper rate. Forcing the associated counter to re-load the scale register value every time PWMSCLA or
PWMSCLB is written prevents this.
Writing to the scale registers while channels are operating can cause irregularities in the PWM outputs.
29.3.1.3
Clock Select
Each PWM channel has the capability of selecting one of two clocks. For channels 0, 1, 4, and 5 the clock choices are clock A or SA. For channels 2, 3, 6 and 7, the choices are clock B or SB. The clock selection is done with the PWMCLK[PCLK x ] control bits.
Changing clock control bits while channels are operating can cause irregularities in the PWM outputs.
29.3.2
PWM Channel Timers
The main part of the PWM module is the actual timers. Each of the timer channels has a counter, a period register, and a duty register (each are 8-bit). The waveform output period is controlled by a match between the period register and the value in the counter. The duty is controlled by a match between the duty register and the counter value and causes the state of the output to change during the period. The starting polarity of the output is also selectable on a per channel basis.
shows a block diagram for a PWM timer.
29-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Pulse Width Modulation (PWM) Module
Clock Source
From
PWMCNT n PWMDTY n 0
1
PWME n
PWMOUT n
PPOL n
PWMPER n
PWMCAE = 1
PWMCAE = 0
Figure 29-15. PWM Timer Channel Block Diagram
29.3.2.1
PWM Enable
Each PWM channel has an enable bit (PWME n ) to start its waveform output. When any of the PWME n bits are set (PWME n =1), the associated PWM output signal is enabled immediately. However, the actual
PWM waveform is not available on the associated PWM output until its clock source begins its next cycle; this is due to the synchronization of PWME n and the clock source. An exception is when channels are concatenated. Refer to
Section 29.3.2.7, “PWM 16-Bit Functions”
for more detail.
Note that the first PWM cycle after enabling the channel can be irregular. When the channel is disabled
(PWME n =0), the counter for the channel does not count.
29.3.2.2
PWM Polarity
Each channel has a polarity bit to allow starting a waveform cycle with a high or low signal. This is shown on the block diagram as a mux select. When one of the bits in the PWMPOL register is set, the associated
PWM channel output is high at the beginning of the waveform, then goes low when the duty count is reached. Conversely, if the polarity bit is zero, the output starts low and then goes high when the duty count is reached.
29.3.2.3
PWM Period and Duty
Dedicated period and duty registers exist for each channel and are double buffered so that if they change while the channel is enabled, the change will not take effect until one of the following occurs:
• The effective period ends
• The PWMCNT n register is written (counter resets to 0x00)
• The channel is disabled, PWME n = 0
In this way, the output of the PWM will always be either the old waveform or the new waveform, not some variation in between. If the channel is not enabled, then writes to the period and duty registers will go directly to the latches as well as the buffer.
A change in duty or period can be forced into effect immediately by writing the new value to the duty and/or period registers and then writing to the counter. This forces the counter to reset and the new duty
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
29-15
Pulse Width Modulation (PWM) Module and/or period values to be latched. In addition, because the counter is readable, it is possible to know where the count is with respect to the duty value, and software can be used to make adjustments . When forcing a new period or duty into effect immediately, an irregular PWM cycle can occur.
Depending on the polarity bit, the duty registers will contain the count of either the high time or the low time.
29.3.2.4
PWM Timer Counters
Each channel has a dedicated 8-bit up/down counter that runs at the rate of the selected clock source (see
for the available clock sources and rates). The counter compares to two registers, a duty register and a period register, as shown in
. When the PWM counter matches the duty register, the output flip-flop changes state, causing the PWM waveform to also change state. A match between the
PWM counter and the period register behaves differently depending on what output mode is selected as shown in
and described in Section 29.3.2.5, “Left-Aligned Outputs”
and
Each channel counter can be read at anytime without affecting the count or the operation of the PWM channel.
Any value written to the counter causes the counter to reset to 0x00, the counter direction to be set to up, the immediate load of both duty and period registers with values from the buffers, and the output to change according to the polarity bit. When the channel is disabled (PWME n = 0), the counter stops. When a channel becomes enabled (PWME n = 1), the associated PWM counter continues from the count in the
PWMCNT n register. This allows the waveform to continue where it left off when the channel is re-enabled. When the channel is disabled, writing “0” to the period register will cause the counter to reset on the next selected clock.
NOTE
If the user wants to start a new clean PWM waveform without any history from the old waveform, the user must write to channel counter
(PWMCNT n ) prior to enabling the PWM channel (PWME n = 1).
Generally, writes to the counter are done prior to enabling a channel in order to start from a known state.
However, writing a counter can also be done while the PWM channel is enabled (counting). The effect is similar to writing the counter when the channel is disabled, except that the new period is started immediately with the output set according to the polarity bit. Writing to the counter while the channel is enabled can cause an irregular PWM cycle to occur.
The counter is cleared at the end of the effective period (see
Section 29.3.2.5, “Left-Aligned Outputs ” and
Section 29.3.2.6, “Center-Aligned Outputs”
for more details).
Table 29-14. PWM Timer Counter Conditions
Counter Clears (0x00)
When PWMCNT n register written to any value
Effective period ends
Counter Counts Counter Stops
When PWM channel is enabled
(PWME n = 1). Counts from last value in PWMCNT n .
When PWM channel is disabled
(PWME n = 0)
29-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Pulse Width Modulation (PWM) Module
29.3.2.5
Left-Aligned Outputs
The PWM timer provides the choice of two types of outputs: left- or center-aligned. They are selected with the PWMCAE[CAE n ] bits. If the CAE n bit is cleared, the corresponding PWM output will be left-aligned.
In left-aligned output mode, the 8-bit counter is configured as an up counter only. It compares to two
PWM counter matches the duty register, the output flip-flop changes state causing the PWM waveform to also change state. A match between the PWM counter and the period register resets the counter and the output flip-flop, as shown in
, as well as performing a load from the double buffer period and
duty register to the associated registers, as described in Figure 29.3.2.3
. The counter counts from 0 to the value in the period register minus 1.
NOTE
Changing the PWM output mode from left-aligned to center-aligned output
(or vice versa) while channels are operating can cause irregularities in the
PWM output. It is recommended to program the output mode before enabling the PWM channel.
PPOL n = 0
PPOL n = 1
PWMDTY n
Period = PWMPER n
Figure 29-16. PWM Left-Aligned Output Waveform
To calculate the output frequency in left-aligned output mode for a particular channel, take the selected clock source frequency for the channel (A, B, SA, or SB) and divide it by the value in the period register for that channel.
PWM n frequency =
PWMPER n
Eqn. 29-7
The PWM n duty cycle (high time as a percentage of period) is expressed as:
Duty Cycle =
⎛
⎝
– n ] –
PWMDTY n
PWMPER n
⎞
⎠
× 100% Eqn. 29-8
29.3.2.5.1
Left-Aligned Output Example
As an example of a left-aligned output, consider the following case:
Clock source = internal bus clock, where internal bus clock = 40 MHz (2.5 ns period)
PPOL n = 0, PWMPER n = 4, PWMDTY n = 1
PWM n frequency = 40 MHz ÷ 4 = 10 MHz
PWM n period = ns
PWMn Duty Cycle = 1 –
4
×
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
29-17
Pulse Width Modulation (PWM) Module
The output waveform generated is below:
E = 2.5ns
DUTY CYCLE = 75%
PERIOD = ns
Figure 29-17. PWM Left-Aligned Output Example Waveform
29.3.2.6
Center-Aligned Outputs
For center-aligned output mode selection, set the PWMCAE[CAE n ] bit and the corresponding PWM output will be center-aligned.
The 8-bit counter operates as an up/down counter in this mode and is set to up whenever the counter is equal to 0x00. The counter compares to two registers, a duty register and a period register, as shown in the
block diagram in Figure 29-15 . When the PWM counter matches the duty register, the output flip-flop
changes state, causing the PWM waveform to also change state. A match between the PWM counter and the period register changes the counter direction from an up-count to a down-count. When the PWM counter decrements and matches the duty register again, the output flip-flop changes state causing the
PWM output to also change state. When the PWM counter decrements and reaches zero, the counter direction changes from a down-count back to an up-count, and a load from the double buffer period and duty registers to the associated registers is performed as described in
from 0 up to the value in the period register and then back down to 0. Thus the effective period is
PWMPER n × 2.
Changing the PWM output mode from left-aligned output to center-aligned output (or vice versa) while channels are operating can cause irregularities in the PWM output. It is recommended to program the output mode before enabling the PWM channel.
PPOL n = 0
PPOL n = 1
PWMDTY n
PWMPER n
PWMDTY n
PWMPER n
PWM n frequency =
Period = PWMPER n
×
2
Figure 29-18. PWM Center-Aligned Output Waveform
To calculate the output frequency in center-aligned output mode for a particular channel, take the selected clock source frequency for the channel (A, B, SA, or SB) and divide it by twice the value in the period register for that channel.
Eqn. 29-9
29-18
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Pulse Width Modulation (PWM) Module
The PWM n duty cycle (high time as a percentage of period) is expressed as:
Duty Cycle = – n ] –
PWMDTY n
PWMPER n
× 100%
29.3.2.6.1
Center-Aligned Output Example
As an example of a center-aligned output, consider the following case:
Clock source = internal bus clock, where internal bus clock = 40 MHz (2.5 ns period)
PPOL n = 0, PWMPER n = 4, PWMDTY n = 1
PWM n frequency = 40 MHz / (2 × 4) = 5 MHz
PWM n period = ns
PWMn Duty Cycle =
⎛
⎝
1 –
4
⎞
⎠
×
Shown below is the generated output waveform.
Eqn. 29-10
E = 2.5ns
E = 2.5ns
DUTY CYCLE = 75%
PERIOD = ns
Figure 29-19. PWM Center-Aligned Output Example Waveform
29.3.2.7
PWM 16-Bit Functions
The PWM timer also has the option of generating 48-bit channels or four 16-bit channels for greater PWM resolution. This 16-bit channel option is achieved through the concatenation of two 8-bit channels.
The PWMCTL register contains four concatenation control bits, each used to concatenate a pair of PWM channels into one 16-bit channel. Channels 0 and 1 are concatenated with the CON01 bit, channels 2 and
3 are concatenated with the CON23 bit, and so on. Change these bits only when both corresponding channels are disabled.
order bytes of the double byte channel. When channels 0 and 1 are concatenated, channel 0 registers become the high order bytes of the double byte channel.
When using the 16-bit concatenated mode, the clock source is determined by the low order 8-bit channel clock select control bits (the odd numbered channel). The resulting PWM is output to the pins of the
corresponding low order 8-bit channel, as shown in Figure 29-20 . The polarity of the resulting PWM
output is controlled by the PPOL n bit of the corresponding low order 8-bit channel as well.
Once concatenated mode is enabled (PWMCTL[CON nn ] bits set), enabling/disabling the corresponding
16-bit PWM channel is controlled by the low order PWME n bit. In this case, the high order bytes’ PWME n bits have no effect, and their corresponding PWM output is disabled.
In concatenated mode, writes to the 16-bit counter by using a 16-bit access or writes to either the low or high order byte of the counter will reset the 16-bit counter. Reads of the 16-bit counter must be made by
16-bit access to maintain data coherency.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
29-19
Pulse Width Modulation (PWM) Module
Clock Source 7 High
PWMCNT6
Low
PWMCNT7
Clock Source 5
Period/Duty Compare
High
PWMCNT4
Low
PWMCNT5
Period/Duty Compare
Clock Source 3 High
PWMCNT2
Low
PWMCNT3
Period/Duty Compare
Clock Source 1 High
PWMCNT0
Low
PWMCNT1
PWMOUT7
PWMOUT5
PWMOUT3
Period/Duty Compare PWMOUT1
Figure 29-20. PWM 16-Bit Mode
Either left- or center-aligned output mode can be used in concatenated mode and is controlled by the low order CAE n bit. The high order CAE n bit has no effect. The table shown below is used to summarize which channels are used to set the various control bits when in 16-bit mode.
Table 29-15. 16-bit Concatenation Mode Summary
CON nn
CON67
CON45
CON23
CON01
PWME n
PWM7
PWM5
PWME3
PWME1
PPOL n
PPOL7
PPOL5
PPOL3
PPOL1
PCLK n
PCLK7
PCLK5
PCLK3
PCLK1
CAE n
CAE7
CAE5
CAE3
CAE1
PWM n
Output
PWMOUT7
PWMOUT5
PWMOUT3
PWMOUT1
29.3.2.8
PWM Boundary Cases
The following table summarizes the boundary conditions for the PWM regardless of the output mode (left- or center-aligned) and 8-bit (normal) or 16-bit (concatenation):
29-20
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Pulse Width Modulation (PWM) Module
Table 29-16. PWM Boundary Cases
PWMDTY n PWMPER n
0x00
(indicates no duty)
>0x00
0x00
(indicates no duty)
>0x00
XX
XX
0x00
1
(indicates no period)
0x00
1
(indicates no period)
≥
PWMPER n
≥
PWMPER n
1
Counter = 0x00 and does not count.
XX
XX
PPOL n
1
0
1
0
1
0
PWM n Output
Always Low
Always High
Always High
Always Low
Always High
Always Low
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
29-21
Pulse Width Modulation (PWM) Module
29-22
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 30
FlexCAN
30.1
Introduction
The FlexCAN is a communication controller implementing the controller area network (CAN) protocol, an asynchronous communications protocol used in automotive and industrial control systems. It is a high speed (1 Mbps), short distance, priority-based protocol that can communicate using a variety of mediums
(such as fiber optic cable or an unshielded twisted pair of wires). The FlexCAN supports both the standard and extended identifier (ID) message formats specified in the CAN protocol specification, revision 2.0, part B.
The CAN protocol was primarily, but not only, designed to be used as a vehicle serial data bus, meeting the specific requirements of this field: real-time processing, reliable operation in the EMI environment of a vehicle, cost-effectiveness, and required bandwidth. A general working knowledge of the CAN protocol revision 2.0 is assumed in this document. For details, refer to the CAN protocol revision 2.0 specification.
30.1.1
Block Diagram
A block diagram describing the various submodules of the FlexCAN module is shown in
.
Each submodule is described in detail in subsequent sections. The message buffer architecture is shown in
.
FlexCAN
MB15
MB14
MB3
MB2
MB1
MB0
••
••
••
••
••
••
••
••
••
••
Message
Buffer
Management
Max MB #
[0:15]
CAN
Protocol
Interface
Bus Interface Unit
CANTX
CANRX
Clocks, Address and Data Buses,
Interrupt and Test Signals
Internal Bus Interface
Figure 30-1. FlexCAN Block Diagram and Pinout
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 30-1
FlexCAN
Tx
Rx
Serial Buffers
Tx Shifter
Rx Shifter
Transparent to User
Mask 14
Mask 15
Global Mask
Control Interrupt Request
Data
•
•
•
16 Transmit/Receive
Message Buffers •
•
•
•
Buffer 13
Buffer 0
Data
Buffer 14
Buffer 15
Data
Data Length
Time Stamp
ID
•
•
•
• •
•
•
Figure 30-2. FlexCAN Message Buffer Architecture
30.1.1.1
The CAN System
A typical CAN system is shown below in
.
CAN Station 1 CAN Station 2 CAN Station n
ColdFire Processor
CANTX
FlexCAN
CANRX
Transceiver
CAN Bus
Figure 30-3. Typical CAN System
Each CAN station is connected physically to the CAN bus through a transceiver. The transceiver provides the transmit drive, waveshaping, and receive/compare functions required for communicating on the CAN bus. It can also provide protection against damage to the FlexCAN caused by a defective CAN bus or defective stations.
30-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
FlexCAN
30.1.2
Features
Following are the main features of the FlexCAN module:
• Full implementation of the CAN protocol specification version 2.0B
— Standard data and remote frames (up to 109 bits long)
— Extended data and remote frames (up to 127 bits long)
— 0–8 bytes data length
— Programmable bit rate up to 1 Mbps
— Content-related addressing
• Up to16 flexible message buffers of zero to eight bytes data length, each configurable as Rx or Tx, all supporting standard and extended messages
• Listen-only mode capability
• Three programmable mask registers: global (for MBs 0–13), special for MB14, and special for
MB15
• Programmable transmission priority scheme: lowest ID or lowest buffer number
• Time stamp based on 16-bit, free-running timer
• Global network time, synchronized by a specific message
• Programmable I/O modes
• Maskable interrupts
• Independent of the transmission medium (an external transceiver is assumed)
• Open network architecture
• Multimaster bus
• High immunity to EMI
• Short latency time due to an arbitration scheme for high-priority messages
30.1.3
Modes of Operation
30.1.3.1
Normal Mode
In normal mode, the module operates receiving and/or transmitting message frames, errors are handled normally, and all the CAN protocol functions are enabled. User and supervisor modes differ in the access to some restricted control registers.
30.1.3.2
Freeze Mode
Freeze mode is entered by setting:
• CANMCR[FRZ], and
• CANMCR[HALT], or by asserting the BKPT signal.
Once entry into freeze mode is requested, the FlexCAN waits until an intermission or idle condition exists on the CAN bus, or until the FlexCAN enters the error passive or bus off state. Once one of these
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 30-3
FlexCAN conditions exists, the FlexCAN waits for the completion of all internal activity such as arbitration, matching, move-in, and move-out. When this happens, the following events occur:
• The FlexCAN stops transmitting/receiving frames.
• The prescaler is disabled, thus halting all CAN bus communication.
• The FlexCAN ignores its Rx pins and drives its Tx pins as recessive.
• The FlexCAN loses synchronization with the CAN bus and the NOTRDY and FRZACK bits in
CANMCR are set.
• The CPU is allowed to read and write the error counter registers (in other modes they are read-only).
After engaging one of the mechanisms to place the FlexCAN in freeze mode, the user must wait for the
FRZACK bit to be set before accessing any other registers in the FlexCAN; otherwise, unpredictable operation may occur. In freeze mode, all memory mapped registers are accessible.
To exit freeze mode, the BKPT line must be negated or the HALT bit in CANMCR must be cleared. Once freeze mode is exited, the FlexCAN will resynchronize with the CAN bus by waiting for 11 consecutive recessive bits before beginning to participate in CAN bus communication.
30.1.3.3
Module Disabled Mode
This mode disables the FlexCAN module; it is entered by setting CANMCR[MDIS]. If the module is disabled during freeze mode, it shuts down the system clocks, sets the LPMACK bit, and clears the
FRZACK bit.
If the module is disabled during transmission or reception, FlexCAN does the following:
• Waits to be in either idle or bus-off state, or else waits for the third bit of intermission and then checks it to be recessive
• Waits for all internal activities such as arbitration, matching, move-in, and move-out to finish
• Ignores its Rx input pin and drives its Tx pin as recessive
• Shuts down the system clocks
The bus interface unit continues to operate, enabling the CPU to access memory-mapped registers, except the free-running timer, the error counter register, and the message buffers, which cannot be accessed when the module is disabled. Exiting from this mode is done by negating the MDIS bit, which will resume the clocks and negate the LPMACK bit.
30.1.3.4
Loop-back Mode
The module enters this mode when the LPB bit in the control register is set. In this mode, FlexCAN performs an internal loop back that can be used for self test operation. The bit stream output of the transmitter is internally fed back to the receiver input. The Rx CAN input pin is ignored and the Tx CAN output goes to the recessive state (logic 1). FlexCAN behaves as it normally does when transmitting and treats its own transmitted message as a message received from a remote node. In this mode, FlexCAN ignores the bit sent during the ACK slot in the CAN frame acknowledge field to ensure proper reception of its own message. Both transmit and receive interrupts are generated.
30-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
FlexCAN
30.1.3.5
Listen-only Mode
In listen-only mode, transmission is disabled, all error counters are frozen and the module operates in a
CAN error passive mode. Only messages acknowledged by another CAN station will be received. If
FlexCAN detects a message that has not been acknowledged, it will flag a BIT0 error (without changing the REC), as if it was trying to acknowledge the message. Because the module does not influence the CAN bus in this mode, the device is capable of functioning like a monitor or for automatic bit-rate detection.
30.2
External Signal Description
Each FlexCAN module has two I/O signals connected to the external MPU pins: CANTX and CANRX.
CANTX transmits serial data to the CAN bus transceiver, while CANRX receives serial data from the
CAN bus transceiver.
30.3
Memory Map/Register Definition
The FlexCAN module address space is split into 128 bytes starting at the base address, and then an extra
256 bytes starting at the base address +128. The upper 256 are fully used for the message buffer structures,
as described in Section 30.3.9, “Message Buffer Structure.”
Out of the lower 128 bytes, only part is occupied by various registers.
Table 30-1. FlexCAN Memory Map
IPSBAR
OffsetAddres s
FlexCAN
Register
Width
(bits)
Affected by Hard
Reset
Affected by Soft
Reset
0x1C_0FC02_
0000
FlexCAN Module Configuration
Register (CANMCR)
Supervisor-only Access Registers
32 Y Y
Supervisor/User Access Registers
32 Y N 0x1C_0FC02_
0004
FlexCAN Control Register
(CANCTRL)
0x1C_0FC02_
0008
Free Running Timer (TIMER)
0xC_0FC02_0
010
Rx Global Mask (RXGMASK)
0x1C_0FC02_
0014
Rx Buffer 14 Mask (RX14MASk)
0x1C_0FC02_
0018
Rx Buffer 15 Mask (RX15MASK)
0x1C_0FC02_
001C
Error Counter Register (ERRCNT)
0x1C_0FC02_
0020
Error and Status Register
(ERRSTAT)
32
32
32
32
32
32
Y
Y
Y
Y
Y
Y
Y
N
N
N
Y
Y
Access Reset Value Section/Page
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
0xD890_000F
0x0000_0000
0x0000_0000
0x1FFF_FFFF
0x1FFF_FFFF
0x1FFF_FFFF
0x0000_0000
0x0000_0000
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 30-5
FlexCAN
Table 30-1. FlexCAN Memory Map (continued)
IPSBAR
OffsetAddres s Register
FlexCAN
0x1C_0FC02_
0028
Interrupt Mask Register (IMASK)
0x1C_0FC02_
0030
Interrupt Flag Register (IFLAG)
0x1C_0FC02_
0080
Message Buffers 0–15 (MB0–15)
Width
(bits)
Affected by Hard
Reset
Affected by Soft
Reset
Access Reset Value Section/Page
32
32
2048
Y
Y
N
Y
Y
N
R/W 0x0000_0000
R/W 0x0000_0000
R/W —
NOTE
The FlexCAN has no hard-wired protection against invalid bit/field programming within its registers. Specifically, no protection is provided if the programming does not meet CAN protocol requirements.
Programming the FlexCAN control registers is typically done during system initialization, prior to the
FlexCAN becoming synchronized with the CAN bus. The configuration registers can be changed after synchronization by halting the FlexCAN module. This is done when the user sets the CANMCR[HALT] bit. The FlexCAN responds by setting the CANMCR[NOTRDY] bit.
30.3.1
FlexCAN Configuration Register (CANMCR)
CANMCR defines global system configurations, such as the module operation mode and maximum message buffer configuration. Most of the fields in this register can be accessed at any time, except the
MAXMB field, which should only be changed while the module is in freeze mode.
Access: Supervisor read/write IPSBAR
OffsetAd dress:
0x1C_0FC02_0000 (CANMCR)
31 30
R
MDIS FRZ
W
Reset 1 1
29
0
0
28
HALT
1
27
NOT
RDY
1
26
0
0
25
SOFT
RST
24 23
FRZ
ACK SUPV
0 0 1
22
0
0
21
0
0
20
LPM
ACK
1
19
0
0
18
0
0
17
0
0
16
0
0
15
R 0
W
Reset 0
14
0
0
13
0
12
0
11
0
10
0
9
0
8
0
7
0
6
0
5
0
4
0
3 2 1
MAXMB
1 1
0
1 0 0 0 0 0 0 0 0 0 0
Figure 30-4. FlexCAN Configuration Register (CANMCR)
1
30-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
FlexCAN
Table 30-2. CANMCR Field Descriptions
Field Description
31
MDIS
30
FRZ
Module disable. This bit controls whether FlexCAN is enabled or not. When disabled, FlexCAN shuts down the
FlexCAN clocks that drive the CAN interface and Message Buffer sub-module. This is the only bit in CANMCR not affected by soft reset. See
Section 30.1.3.3, “Module Disabled Mode,”
for more information.
0 Enable the FlexCAN module, clocks enabled
1 Disable the FlexCAN module, clocks disabled
for more information.
0 FlexCAN ignores the BKPT signal and the CANMCR[HALT] bit.
1 FlexCAN module enabled to enter debug mode.
Reserved, should be cleared.
29
28
HALT
Halt FlexCAN. Setting this bit puts the FlexCAN module into freeze mode. It has the same effect as assertion of the
BKPT signal. This bit is set after reset and should be cleared after initializing the message buffers and control registers. FlexCAN message buffer receive and transmit functions are inactive until this bit is cleared. While in
0 The FlexCAN operates normally
1 FlexCAN enters freeze mode if FRZ = 1
27
NOTRDY
FlexCAN not ready. This bit indicates that the FlexCAN is either in disable or freeze mode. This bit is read-only and it is cleared once the FlexCAN exits these modes.
0 FlexCAN is either in normal mode, listen-only mode, or loop-back mode.
h1FlexCAN is in disable or freeze mode.
26 Reserved, should be cleared.
25
SOFTRST
Soft reset. When set, the FlexCAN resets its internal state machines (sequencer, error counters, error flags, and timer) and the host interface registers (CANMCR [except the MDIS bit], TIMER, ERRCNT, ERRSTAT, IMASK, and
IFLAG).
The configuration registers that control the interface with the CAN bus are not changed (CANCTRL, RXGMASK,
RX14MASK, RX15MASK). Message buffers are also not changed. This allows SOFTRST to be used as a debug feature while the system is running.
Since soft reset is synchronous and has to follow a request/acknowledge procedure across clock domains, it may take some time to fully propagate its effect. The SOFTRST bit remains set while reset is pending and is automatically cleared when reset completes. The user should poll this bit to know when the soft reset has completed.
0 Soft reset cycle completed
1 Soft reset cycle initiated
24
FRZACK
Freeze acknowledge. Indicates that the FlexCAN module has entered freeze mode. The user should poll this bit after freeze mode has been requested, to know when the module has actually entered freeze mode. When freeze mode is exited, this bit is cleared once the FlexCAN prescaler is enabled. This is a read-only bit.
0 The FlexCAN has exited freeze mode and the prescaler is enabled.
1 The FlexCAN has entered freeze mode, and the prescaler is disabled.
23
SUPV
Supervisor/user data space. Places the FlexCAN registers in either supervisor or user data space.
0 Registers with access controlled by the SUPV bit are accessible in either user or supervisor privilege mode.
1 Registers with access controlled by the SUPV bit are restricted to supervisor mode.
22–21 Reserved, should be cleared.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 30-7
FlexCAN
Table 30-2. CANMCR Field Descriptions (continued)
Field Description
20
LPMACK
Low power mode acknowledge. Indicates that FlexCAN is disabled. Disabled mode cannot be entered until all current transmission or reception processes have finished, so the CPU can poll the LPMACK bit to know when the
FlexCAN has actually entered low power mode. See Section 30.1.3.3, “Module Disabled Mode,”
and Chapter 8,
“Power Management,” for more information. This bit is read-only.
0 FlexCAN not disabled.
1 FlexCAN is in disabled mode.
19–4
3–0
MAXMB
Reserved, should be cleared.
Maximum number of message buffers. Defines the maximum number of message buffers that will take part in the matching and arbitration process. The reset value (0xF) is equivalent to16 message buffer (MB) configuration. This field should be changed only while the module is in freeze mode.
Note:
Maximum MBs in Use = MAXMB + 1
30.3.2
FlexCAN Control Register (CANCTRL)
CANCTRL is defined for specific FlexCAN control features related to the CAN bus, such as bit-rate, programmable sampling point within an Rx bit, loop back mode, listen-only mode, bus off recovery behavior, and interrupt enabling. It also determines the division factor for the clock prescaler. Most of the fields in this register should only be changed while the module is disabled or in freeze mode. Exceptions are the BOFFMSK, ERRMSK, and BOFFREC bits, which can be accessed at any time.
Access: User read/write IPSBAR
OffsetAd dress:
0x1C_0FC02_0004 (CANCTRL)
30 29 28 27 31
R
W
Reset 0 0 0
PRESDIV
0 0
26
0
25
0
24
0
23
RJW
22
0 0
21 20
PSEG1
19
0 0 0
18
0
17
PSEG2
0
16
0
15
R BOFF
W MSK
Reset 0
14
ERR
MSK
0
13
CLK_
SRC
0
12
LPB
0
11
0
10
0
9
0
8
0
7
SMP
6
BOFF
REC
0
5
TSYN LBUF LOM
0
4
0
3
0 0 0 0 0 0
Figure 30-5. FlexCAN Control Register (CANCTRL)
2 1
PROPSEG
0
0 0 0
30-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
FlexCAN
Table 30-3. CANCTRL Field Descriptions
Field Description
31–24
PRESDIV
Prescaler division factor. Defines the ratio between the clock source frequency (set by CLK_SRC bit) and the serial clock (S clock) frequency. The S clock period defines the time quantum of the CAN protocol. For the reset value, the S clock frequency is equal to the clock source frequency. The maximum value of this register is 0xFF, that gives a minimum S clock frequency equal to the clock source frequency divided by 256. For more information refer to
.”
S clock frequency =
S clock frequency =
Eqn. 30-1
Eqn. 30-2
23–22
RJW
21–19
PSEG1
18–16
PSEG2
Resynchronization jump width. Defines the maximum number of time quanta (one time quantum is equal to the S clock period) that a bit time can be changed by one resynchronization. The valid programmable values are 0 – 3.
Resync jump width = (RJW + 1) time quanta
Phase buffer segment 1. Defines the length of phase buffer segment 1 in the bit time. The valid programmable values are 0 – 7.
Phase buffer segment 1 = (PSEG1 + 1) time quanta
Phase buffer segment 2. Defines the length of phase buffer segment 2 in the bit time. The valid programmable values are 1 – 7.
Phase buffer segment 2 = (PSEG2 + 1) time quanta
15
BOFFMSK
Bus off interrupt mask.
0 Bus off interrupt disabled
1 Bus off interrupt enabled
14
ERRMSK
Error interrupt mask.
0 Error interrupt disabled
1 Error interrupt enabled
13
CLK_SRC
Clock source. Selects the clock source for the CAN interface to be fed to the prescalar. This bit should only be changed while the module is disabled.
0 Clock source is EXTAL
1 Clock source is the internal bus clock, f sys/23
12
LPB
Loop back. Configures FlexCAN to operate in loop-back mode. In this mode, FlexCAN performs an internal loop back that can be used for self test operation. The bit stream output of the transmitter is fed back internally to the receiver input. The Rx CAN input pin is ignored and the Tx CAN output goes to the recessive state (logic 1).
FlexCAN behaves as it normally does when transmitting, and treats its own transmitted message as a message received from a remote node. In this mode, FlexCAN ignores the bit sent during the ACK slot in the CAN frame acknowledge field, generating an internal acknowledge bit to ensure proper reception of its own message. Both transmit and receive interrupts are generated.
0 Loop back disabled
1 Loop back enabled
11–8
7
SMP
Reserved, should be cleared.
Sampling mode. Determines whether the FlexCAN module will sample each received bit one time or three times to determine its value.
0 One sample, taken at the end of phase buffer segment 1, is used to determine the value of the received bit.
1 Three samples are used to determine the value of the received bit. The samples are taken at the normal sample point and at the two preceding periods of the S-clock; a majority rule is used.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 30-9
FlexCAN
Table 30-3. CANCTRL Field Descriptions (continued)
Field Description
6
BOFFREC
Bus off recovery mode. Defines how FlexCAN recovers from bus off state. If this bit is cleared, automatic recovering from bus off state occurs according to the CAN Specification 2.0B
. If the bit is set, automatic recovering from bus off is disabled and the module remains in bus off state until the bit is cleared by the user. If the bit is cleared before
128 sequences of 11 recessive bits are detected on the CAN bus, then bus off recovery happens as if the
BOFFREC bit had never been set. If the bit is cleared after 128 sequences of 11 recessive bits occurred, then
FlexCAN will re-synchronize to the bus by waiting for 11 recessive bits before joining the bus. After clearing, the
BOFFREC bit can be set again during bus off, but it will only be effective the next time the module enters bus off.
If BOFFREC was cleared when the module entered bus off, setting it during bus off will not be effective for the current bus off recovery.
0 Automatic recovering from bus off state enabled, according to CAN Spec 2.0B
1 Automatic recovering from bus off state disabled
5
TSYN
Timer synchronize mode. Enables the mechanism that resets the free-running timer each time a message is received in Message Buffer 0. This feature provides the means to synchronize multiple FlexCAN stations with a special “SYNC” message (global network time).
0 Timer synchronization disabled.
1 Timer synchronization enabled.
Note: There can be a bit clock skew of four to five counts between different FlexCAN modules that are using this feature on the same network.
4
LBUF
Lowest buffer transmitted first. Defines the ordering mechanism for message buffer transmission.
0 Message buffer with lowest ID is transmitted first
1 Lowest numbered buffer is transmitted first
3
LOM
Listen-only mode. Configures FlexCAN to operate in listen-only mode. In this mode transmission is disabled, all error counters are frozen, and the module operates in a CAN error passive mode. Only messages acknowledged by another CAN station will be received. If FlexCAN detects a message that has not been acknowledged, it will flag a BIT0 error (without changing the REC), as if it was trying to acknowledge the message.
0 FlexCAN module is in normal active operation; listen-only mode is deactivated
1 FlexCAN module is in listen-only mode operation
2–0
PROPSEG
Propagation segment. Defines the length of the propagation segment in the bit time. The valid programmable values are 0 – 7.
Propagation segment time = (PROPSEG + 1) time-quanta
Note: A time-quantum = 1 S clock period.
30.3.3
FlexCAN Free Running Timer Register (TIMER)
This register represents a 16-bit free running counter that can be read and written to by the CPU. The timer starts from 0x0000 after reset, counts linearly to 0xFFFF, and wraps around.
The timer is clocked by the FlexCAN bit-clock (which defines the baud rate on the CAN bus). During a message transmission/reception, it increments by one for each bit that is received or transmitted. When there is no message on the bus, it counts using the previously programmed baud rate. During freeze mode, the timer is not incremented.
The timer value is captured at the beginning of the identifier (ID) field of any frame on the CAN bus. This captured value is written into the TIMESTAMP entry in a message buffer after a successful reception or transmission of a message.
Writing to the timer is an indirect operation. The data is first written to an auxiliary register, then an internal request/acknowledge procedure across clock domains is executed. All this is transparent to the user, except
30-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
FlexCAN for the fact that the data will take some time to be actually written to the register. If desired, software can poll the register to discover when the data was actually written.
IPSBAR
OffsetAd dress:
0x1C_0FC02_0008 (TIMER) Access: User read/write
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
R 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TIMER
W
Reset 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Figure 30-6. FlexCAN Timer Register (TIMER)
Table 30-4. TIMER Field Descriptions
Field Description
31–16 Reserved, should be cleared.
15–0
TIMER
Free running timer. Captured at the beginning of the identifier (ID) field of any frame on the CAN bus. This captured value is written into the TIMESTAMP entry in a message buffer after a successful reception or transmission of a message.
30.3.4
Rx Mask Registers (RXGMASK, RX14MASK, RX15MASK)
These registers are used as acceptance masks for received frame IDs. Three masks are defined: a global mask (RXGMASK) used for Rx buffers 0–13 and two separate masks for buffers 14 (RX14MASK) and
15 (RX15MASK). The meaning of each mask bit is the following:
MI n bit = 0: The corresponding incoming ID bit is “don’t care”.
MI n bit = 1: The corresponding ID bit is checked against the incoming ID bit, to see if a match exists.
Note that these masks are used both for standard and extended ID formats. The value of the mask registers should not be changed while in normal operation (only while in freeze mode), as locked frames that matched a message buffer (MB) through a mask may be transferred into the MB (upon release) but may no longer match.
Table 30-5. Mask Examples for Normal/Extended Messages
IDE
Extended ID
ID17......................................ID0
Match
MB2-ID
MB3-ID
MB4-ID
MB5-ID
MB14-ID
Rx_Global_Mask
Rx_Msg in
1
Base ID
ID28.................ID18
1 1 1 1 1 1 1 1 0 0 0
1 1 1 1 1 1 1 1 0 0 0
0 0 0 0 0 0 1 1 1 1 1
0 0 0 0 0 0 1 1 1 0 1
1 1 1 1 1 1 1 1 0 0 0
1 1 1 1 1 1 1 1 1 1 0
1 1 1 1 1 1 1 1 0 0 1
0
1
0
1
1
1
0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 1
0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 MB3
1
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 30-11
FlexCAN
Table 30-5. Mask Examples for Normal/Extended Messages (continued)
Base ID
ID28.................ID18
IDE
Extended ID
ID17......................................ID0
Rx_Msg in
2
Rx_Msg in
3
Rx_Msg in
4
Rx_Msg in
5
1 1 1 1 1 1 1 1 0 0 1
1 1 1 1 1 1 1 1 0 0 1
0 1 1 1 1 1 1 1 0 0 0
0 1 1 1 1 1 1 1 0 0 0
0
1
0
1
0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 0
0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
RX14MASK
Rx_Msg in
6
0 1 1 1 1 1 1 1 1 1 1
1 0 1 1 1 1 1 1 0 0 0 1
1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0
0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
Rx_Msg in
7
0 1 1 1 1 1 1 1 0 0 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
2
3
4
5
6
7
1
Match for Extended Format (MB3).
Match for Normal Format. (MB2).
Mismatch for MB3 because of ID0.
Mismatch for MB2 because of ID28.
Mismatch for MB3 because of ID28, Match for MB14 (Uses RX14MASK).
Mismatch for MB14 because of ID27 (Uses RX14MASK).
Match for MB14 (Uses RX14MASK).
Match
MB2
2
3
4
MB14
5
6
MB14
7
OffsetAd dress:
0x1C_0FC02_0010 (RXGMASK)
0x1C_0FC02_0014 (RX14MASK)
0x1C_0FC02_0018 (RX15MASK)
Access: User read/write
R 0 0 0
W
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
MI
Standard ID
MI
Extended ID
Reset 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
Figure 30-7. FlexCAN Rx Mask Registers (RXGMASK, RX14MASK, RX15MASK)
Table 30-6. RX xx MASK Field Descriptions
Field Description
31–29 Reserved, should be cleared.
28–18
MI28–18
Standard ID mask bits. These bits are the same mask bits for the Standard and Extended Formats.
17–0
MI17–0
Extended ID mask bits. These bits are used to mask comparison only in Extended Format.
30.3.5
FlexCAN Error Counter Register (ERRCNT)
This register has two 8-bit fields reflecting the value of two FlexCAN error counters: transmit error counter
(TXECTR) and receive error counter (RXECTR). The rules for increasing and decreasing these counters are described in the CAN protocol and are completely implemented in the FlexCAN module. Both counters are read-only, except in freeze mode, where they can be written by the CPU.
30-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
FlexCAN
Writing to the ERRCNT register while in freeze mode is an indirect operation. The data is first written to an auxiliary register, then an internal request/acknowledge procedure across clock domains is executed.
All this is transparent to the user, except for the fact that the data will take some time to be actually written to the register. If desired, software can poll the register to discover when the data was actually written.
FlexCAN responds to any bus state as described in the protocol, e.g. transmit error-active or error-passive flag, delay its transmission start time (error-passive), and avoid any influence on the bus when in bus off state. The following are the basic rules for FlexCAN bus state transitions:
• If the value of TXECTR or RXECTR increases to be greater than or equal to 128, the FLTCONF field in the error and status register (ERRSTAT) is updated to reflect error-passive state.
• If the FlexCAN state is error-passive, and either TXECTR or RXECTR decrements to a value less than or equal to 127 while the other already satisfies this condition, the ERRSTAT[FLTCONF] field is updated to reflect error-active state.
• If the value of TXECTR increases to be greater than 255, the ERRSTAT[FLTCONF] field is updated to reflect bus off state, and an interrupt may be issued. The value of TXECTR is then reset to zero.
• If FlexCAN is in bus off state, then TXECTR is cascaded together with another internal counter to count the 128th occurrences of 11 consecutive recessive bits on the bus. Hence, TXECTR is reset to zero and counts in a manner where the internal counter counts 11 such bits and then wraps around while incrementing the TXECTR. When TXECTR reaches the value of 128, the
ERRSTAT[FLTCONF] field is updated to be error-active, and both error counters are reset to zero.
At any instance of a dominant bit following a stream of less than 11 consecutive recessive bits, the internal counter resets itself to zero without affecting the TXECTR value.
• If during system start-up, only one node is operating, then its TXECTR increases in each message it is trying to transmit, as a result of acknowledge errors (indicated by the ERRSTAT[ACKERR] bit). After the transition to error-passive state, the TXECTR does not increment anymore by acknowledge errors. Therefore, the device never goes to the bus off state.
• If the RXECTR increases to a value greater than 127, it is not incremented further, even if more errors are detected while being a receiver. At the next successful message reception, the counter is set to a value between 119 and 127 to resume to error-active state.
IPSBAR
OffsetAd dress:
0x1C_0FC02_001C (ERRCNT) Access: User read/write
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
R 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RXECTR TXECTR
W
Reset 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Figure 30-8. FlexCAN Error Counter Register (ERRCNT)
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 30-13
FlexCAN
Table 30-7. ERRCNT Field Descriptions
Field Description
31–16 Reserved, should be cleared.
15–8
RXECTR
Receive error counter. Indicates current number of receive errors.
7–0
TXECTR
Transmit error counter. Indicates current number of transmit errors.
30.3.6
FlexCAN Error and Status Register (ERRSTAT)
ERRSTAT reflects various error conditions, some general status of the device, and is the source of three interrupts to the CPU. The reported error conditions (bits 15:10) are those occurred since the last time the
CPU read this register. The read action clears bits 15-10. Bits 9–3 are status bits.
Most bits in this register are read only, except for BOFFINT and ERRINT, which are interrupt flags that
can be cleared by writing 1 to them. Writing 0 has no effect. Refer to Section 30.5.1, “Interrupts.”
Access: User read/write IPSBAR
OffsetAd dress:
0x1C_0FC02_0020 (CANCTRL)
31
R 0
W
Reset 0
30
0
0
29
0
0
28
0
0
27
0
0
26
0
0
25
0
0
24
0
0
23
0
0
22
0
0
21
0
0
20
0
0
19
0
0
18
0
0
17
0
0
16
0
0
15
R BIT1
ERR
W
Reset 0
14
BIT0
ERR
13
ACK
ERR
0 0
12
CRC
ERR
11
FRM
ERR
10
STF
ERR
9
TX
WRN
8
RX
WRN
7 6
IDLE TXRX
5
FLT
CONF
4
0 0 0 0 0 0 0 0 0
Figure 30-9. FlexCAN Error and Status Register (ERRSTAT)
3
0
0
2
BOFF
INT w1c
0
1
ERR
INT w1c
0
0
0
0
Table 30-8. ERRSTAT Field Descriptions
Field Description
31–16 Reserved, should be cleared.
15
BIT1ERR
Bit1 error. Indicates inconsistency between the transmitted and received bit in a message.
0 No transmit bit error
1 At least one bit sent as recessive was received as dominant
Note: The transmit bit error field is not modified during the arbitration field or the ACK slot bit time of a message, or by a transmitter that detects dominant bits while sending a passive error frame.
14
BIT0ERR
Bit0 error. Indicates inconsistency between the transmitted and received bit in a message.
0 No transmit bit error
1 At least one bit sent as dominant was received as recessive
30-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
FlexCAN
Table 30-8. ERRSTAT Field Descriptions (continued)
Field Description
13
ACKERR
Acknowledge error. Indicates whether an acknowledgment has been correctly received for a transmitted message.
0 No ACK error was detected since the last read of this register.
1 An ACK error was detected since the last read of this register.
12
CRCERR
Cyclic redundancy check error. Indicates whether or not a CRC error has been detected by the receiver.
0 No CRC error was detected since the last read of this register.
1 A CRC error was detected since the last read of this register.
11
FRMERR
Message form error. Indicates that a form error has been detected by the receiver node, i.e. a fixed-form bit field contains at least one illegal bit.
0 No form error was detected since the last read of this register.
1 A form error was detected since the last read of this register.
10
STFERR
Bit stuff error.
0 No bit stuffing error was detected since the last read of this register.
1 A bit stuffing error was detected since the last read of this register.
9
TXWRN
8
RXWRN
7
IDLE
Transmit error status flag. Reflects the status of the FlexCAN transmit error counter.
0 Transmit error counter
<
96
1 TXErrCounter
≥
96
Receiver error status flag. Reflects the status of the FlexCAN receive error counter.
0 Receive error counter
<
96
1 RxErrCounter
≥
96
Idle status. Indicates when there is activity on the CAN bus.
0 The CAN bus is not idle.
1 The CAN bus is idle.
6
TXRX
Transmit/receive status. Indicates when the FlexCAN module is transmitting or receiving a message. TXRX has no meaning when IDLE = 1.
0 The FlexCAN is receiving a message if IDLE = 0.
1 The FlexCAN is transmitting a message if IDLE = 0.
5–4
FLTCONF
Fault confinement state. Indicates the confinement state of the FlexCAN module, as shown below. If the
CANCTRL[LOM] bit is set, FLTCONF will indicate error-passive. Since the CANCTRL register is not affected by soft reset, the FLTCONF field will not be affected by soft reset if the LOM bit is set.
00 Error active
01 Error passive
1x Bus off
3 Reserved, should be cleared.
2
BOFFINT
Bus off interrupt. Used to request an interrupt when the FlexCAN enters the bus off state. The user must write a 1 to clear this bit. Writing 0 has no effect.
0 No bus off interrupt requested.
1 This bit is set when the FlexCAN state changes to bus off. If the CANCTRL[BOFFMSK] bit is set an interrupt request is generated. This interrupt is not requested after reset.
1
ERRINT
0
Error interrupt. Indicates that at least one of the ERRSTAT[15:10] bits is set. The user must write a 1 to clear this bit. Writing 0 has no effect.
0 No error interrupt request.
1 At least one of the error bits is set. If the CANCTRL[ERRMSK] bit is set, an interrupt request is generated.
Reserved, should be cleared.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 30-15
FlexCAN
30.3.7
Interrupt Mask Register (IMASK)
IMASK contains one interrupt mask bit per buffer. It enables the CPU to determine which buffer will generate an interrupt after a successful transmission/reception (that is, when the corresponding IFLAG bit is set).
IPSBAR
OffsetAd dress:
0x1C_0FC02_0028 (IMASK) Access: User read/write
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
R 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
BUF n M
W
Reset 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Figure 30-10. FlexCAN Interrupt Mask Register (IMASK)
Table 30-9. IMASK Field Descriptions
Field Description
31–16 Reserved, should be cleared.
15–0
BUF n M
Buffer interrupt mask. Enables the respective FlexCAN message buffer (MB0 to MB15) interrupt. These bits allow the CPU to designate which buffers will generate interrupts after successful transmission/reception.
0 The interrupt for the corresponding buffer is disabled.
1 The interrupt for the corresponding buffer is enabled.
Note: Setting or clearing an IMASK bit can assert or negate an interrupt request, if the corresponding IFLAG bit it is set.
30.3.8
Interrupt Flag Register (IFLAG)
IFLAG contains one interrupt flag bit per buffer. Each successful transmission/reception sets the corresponding IFLAG bit and, if the corresponding IMASK bit is set, will generate an interrupt.
The interrupt flag is cleared by writing a 1, while writing 0 has no effect.
IPSBAR
OffsetAd dress:
0x1C_0FC02_0030 (IFLAG) Access: User read/write
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
R 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 BUF n I
W w1c
Reset 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Figure 30-11. FlexCAN Interrupt Flags Register (IFLAG)
30-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
FlexCAN
Table 30-10. IFLAG Field Descriptions
Field Description
31–16 Reserved, should be cleared.
15–0
BUF n I
Buffer interrupt flag. Indicates a successful transmission/reception for the corresponding message buffer. If the corresponding IMASK bit is set, an interrupt request will be generated. The user must write a 1 to clear an interrupt flag; writing 0 has no effect.
0 No such occurrence.
1 The corresponding buffer has successfully completed transmission or reception.
30.3.9
Message Buffer Structure
The message buffer memory map starts at an offset of 0x80 from the FlexCAN’s base address
(0x1C_0FC02_0000). The 256-byte message buffer space is fully used by the 16 message buffer structures.
Each message buffer consists of a control and status field that configures the message buffer, an identifier field for frame identification, and up to 8 bytes of data.
FlexCAN Base
Address Offset
0x80
0x84
0x88
0x8F
0x90
Control/Status
Identifier
8 byte Data fields
Message Buffer 0
Message Buffer 1
0x9F
0xA0
Message Buffer 2
0xAF
0xB0
Message Buffer 3
0x16F
0x170
Message Buffer 14
Message Buffer 15
0x17F
Figure 30-12. FlexCAN Message Buffer Memory Map
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 30-17
FlexCAN
The message buffer structure used by the FlexCAN module is shown in Figure 30-13 . Both standard and
extended frames used in the CAN Specification Version 2.0, Part B are represented. A standard frame is represented by the 11-bit standard identifier, and an extended frame is represented by the combined 29-bits of the standard identifier (11 bits) and the extended identifier (18 bits).
0x0
0x4
31 30 29 28 27 26 25 24 23
CODE
22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SRR IDE RTR LENGTH TIME STAMP
Standard ID[28:18] Extended ID[17:0]
0x8
0xC
Data Byte 0
Data Byte 4
Data Byte 1
Data Byte 5
Data Byte 2
Data Byte 6
Data Byte 3
Data Byte 7
Figure 30-13. Message Buffer Structure for Both Extended and Standard Frames
Table 30-11. Message Buffer Field Descriptions
Field Description
31–28 Reserved, should be cleared.
27–24
CODE
Message buffer code. Can be accessed (read or write) by the CPU and by the FlexCAN module itself, as part of the message buffer matching and arbitration process. The encoding is shown in
Section 30.4, “Functional Overview,”
for additional information.
. See
23
22
SRR
21
IDE
Reserved, should be cleared.
Substitute remote request. Fixed recessive bit, used only in extended format. It must be set by the user for transmission (Tx Buffers) and will be stored with the value received on the CAN bus for Rx receiving buffers. It can be received as either recessive or dominant. If FlexCAN receives this bit as dominant, then it is interpreted as arbitration loss.
0 Dominant is not a valid value for transmission in Extended Format frames
1 Recessive value is compulsory for transmission in Extended Format frames
ID extended bit. Identifies whether the frame format is standard or extended.
0 Standard frame format
1 Extended frame format
20
RTR
19–16
LENGTH
Remote transmission request. Used for requesting transmissions of a data frame. If FlexCAN transmits this bit as
1 (recessive) and receives it as 0 (dominant), it is interpreted as arbitration loss. If this bit is transmitted as 0
(dominant), then if it is received as 1 (recessive), the FlexCAN module treats it as bit error. If the value received matches the value transmitted, it is considered as a successful bit transmission.
0 Indicates the current MB has a data frame to be transmitted
1 Indicates the current MB has a remote frame to be transmitted
Length of data in bytes. Indicates the length (in bytes) of the Rx or Tx data; data is located in offset 0x8 through
0xF of the MB space (see
). In reception, this field is written by the FlexCAN module, copied from the
DLC (data length code) field of the received frame. DLC is defined by the CAN Specification and refers to the data length of the actual frame before it is copied into the message buffer. In transmission, this field is written by the CPU and is used as the DLC field value of the frame to be transmitted.
When RTR is set, the frame to be transmitted is a remote frame and will be transmitted without the DATA field, regardless of the LENGTH field.
15–-0
TIME
STAMP
Free-running counter time stamp. Stores the value of the free-running timer which is captured when the beginning of the identifier (ID) field appears on the CAN bus.
31–29 Reserved, should be cleared.
30-18
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
FlexCAN
Table 30-11. Message Buffer Field Descriptions (continued)
Field Description
28–0
ID
Standard frame identifier: In standard frame format, only the 11 most significant bits (28 to 18) are used for frame identification in both receive and transmit cases. The 18 least significant bits are ignored.
Extended frame identifier: In extended frame format, all bits (both the 11 bits of the standard frame identifier and the 18 bits of the extended frame identifier) are used for frame identification in both receive and transmit cases.
31–24,
23–16,
15–8, 7–0
DATA
Data field. Up to eight bytes can be used for a data frame. For Rx frames, the data is stored as it is received from the CAN bus. For Tx frames, the CPU provides the data to be transmitted within the frame.
Table 30-12. Message Buffer Code for Rx Buffers
Rx Code
BEFORE
Rx New Frame
Description
Rx Code
AFTER
Rx New Frame
Comment
0000
0100
0010
INACTIVE: MB is not active.
EMPTY: MB is active and empty.
FULL: MB is full.
—
0010
0010
MB does not participate in the matching process.
MB participates in the matching process. When a frame is received successfully, the code is automatically updated to FULL.
The act of reading the control & status (C/S) word followed by unlocking the MB does not make the code return to EMPTY. It remains FULL. If a new frame is written to the MB after the C/S word was read and the MB was unlocked, the code still remains
FULL.
0110 OVERRUN: A frame was overwritten into a full buffer.
0110
0010
If the MB is FULL and a new frame should be written into this MB before the CPU had time to read it, the
MB is overwritten, and the code is automatically updated to OVERRUN.
If the code indicates OVERRUN but the CPU reads the C/S word and then unlocks the MB, when a new frame is written to the MB, the code returns to FULL.
0110 If the code already indicates OVERRUN, and yet another new frame must be written, the MB will be overwritten again, and the code will remain
OVERRUN.
0XY1
1
BUSY: Flexcan is updating the contents of the MB with a new receive frame.
The CPU should not try to access the MB.
0010
0110
An EMPTY buffer was written with a new frame (XY was 01).
A FULL/OVERRUN buffer was overwritten (XY was
11).
1
Note that for transmit message buffers (see Table 30-13
), the BUSY bit should be ignored upon read.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 30-19
FlexCAN
MB n [RTR]
X
0
1
0
0
Initial Tx
Code
1000
1100
1100
1010
1110
Table 30-13. Message Buffer Code for Tx Buffers
Code After
Successful
Transmission
—
Description
1000
0100
1010
1010
INACTIVE: Message buffer not ready for transmit and will participate in the arbitration process.
Data frame to be transmitted once, unconditionally. After transmission, the MB automatically returns to the INACTIVE state.
Remote frame to be transmitted unconditionally once, and message buffer becomes an Rx message buffer with the same
ID for data frames.
Transmit a data frame whenever a remote request frame with the same ID is received. This message buffer participates simultaneously in both the matching and arbitration processes.
The matching process compares the ID of the incoming remote request frame with the ID of the MB. If a match occurs, this message buffer is allowed to participate in the current arbitration process and the CODE field is automatically updated to 1110 to allow the MB to participate in future arbitration runs. When the frame is eventually transmitted successfully, the code automatically returns to 1010 to restart the process again.
This is an intermediate code that is automatically written to the message buffer as a result of match to a remote request frame.
The data frame will be transmitted unconditionally once, and then the code will automatically return to 1010. The CPU can also write this code with the same effect.
30.4
Functional Overview
The FlexCAN module is flexible in that each one of its 16 message buffers (MBs) can be assigned either as a transmit buffer or a receive buffer. Each MB, which is up to 8 bytes long, is also assigned an interrupt flag bit that indicates successful completion of either transmission or reception.
An arbitration algorithm decides the prioritization of MBs to be transmitted based on either the message
ID or the MB ordering. A matching algorithm makes it possible to store received frames only into MBs that have the same ID programmed on its ID field. A masking scheme makes it possible to match the ID programmed on the MB with a range of IDs on received CAN frames. Data coherency mechanisms are implemented to guarantee data integrity during MB manipulation by the CPU.
Before proceeding with the functional description, an important concept must be explained. A message buffer is said to be active at a given time if it can participate in the matching and arbitration algorithms that
are happening at that time. An Rx MB with a 0000 code is inactive (refer to Table 30-12
). Similarly, a Tx
MB with a 1000 code is inactive (refer to
Table 30-13 ). An MB not programmed with either 0000 or 1000
will be temporarily deactivated (will not participate in the current arbitration/matching run) when the CPU writes to the C/S field of that MB.
30-20
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
FlexCAN
30.4.1
Transmit Process
The CPU prepares or changes an MB for transmission by writing the following:
1. Control/status word to hold Tx MB inactive (CODE = 1000)
2. ID word
3. Data bytes
4. Control/status word (active CODE, LENGTH)
NOTE
The first and last steps are mandatory.
The first write to the control/status word is important in case there was pending reception or transmission.
The write operation immediately deactivates the MB, removing it from any currently ongoing arbitration or ID matching processes, giving time for the CPU to program the rest of the MB (see
). Once the MB is activated in the fourth step, it will participate in the arbitration process and eventually be transmitted according to its priority. At the end of the successful transmission, the value of the free running timer (TIMER) is written into the message buffer’s time stamp field, the code field in the control and status word is updated, a status flag is set in the IFLAG register, and an interrupt is generated if allowed by the corresponding IMASK register bit. The new code field after transmission depends on the code that was used to activate the MB in step four (see
30.4.2
Arbitration Process
The arbitration process is an algorithm executed by the message buffer management (MBM) that scans the entire MB memory looking for the highest priority message to be transmitted. All MBs programmed as transmit buffers will be scanned to find the lowest ID or the lowest MB number, depending on the
CANCTRL[LBUF] bit.
NOTE
If CANCTRL[LBUF] is cleared, the arbitration considers not only the ID, but also the RTR and IDE bits placed inside the ID at the same positions they are transmitted in the CAN frame.
The arbitration process is triggered in the following events:
• During the CRC field of the CAN frame
• During the error delimiter field of the CAN frame
• During intermission, if the winner MB defined in a previous arbitration was deactivated, or if there was no MB to transmit, but the CPU wrote to the C/S word of any MB after the previous arbitration finished
• When MBM is in idle or bus off state and the CPU writes to the C/S word of any MB
• Upon leaving freeze mode
Once the highest priority MB is selected, it is transferred to a temporary storage space called serial message buffer (SMB), which has the same structure as a normal MB but is not user accessible. This operation is called ‘move-out.’ At the first opportunity window on the CAN bus, the message on the SMB
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 30-21
FlexCAN is transmitted according to the CAN protocol rules. FlexCAN transmits up to 8 data bytes, even if the data
length code (DLC) value is bigger. Refer to Section 30.4.5.1, “Serial Message Buffers (SMBs),”
for more information on serial message buffers.
30.4.3
Receive Process
The CPU prepares or changes an MB for frame reception by writing the following:
1. Control/status word to hold Rx MB inactive (CODE = 0000)
2. ID word
3. Control/status word to mark the Rx MB as active and empty (CODE = 1000)
NOTE
The first and last steps are mandatory.
The first write to the control/status word is important in case there was a pending reception or transmission.
The write operation immediately deactivates the MB, removing it from any currently ongoing arbitration or matching process, giving time for the CPU to program the rest of the MB. Once the MB is activated in the third step, it will be able to receive CAN frames that match the programmed ID. At the end of a successful reception, the value of the free running timer (TIMER) is written into the time stamp field, the received ID, data (8 bytes at most) and length fields are stored, the CODE field in the control and status word is updated (see
), and a status flag is set in the IFLAG register and an interrupt is generated if allowed by the corresponding IMASK bit.
The CPU should read a receive frame from its MB by reading the following:
1. Control/status word (mandatory—activates internal lock for this buffer)
2. ID (optional—needed only if a mask was used)
3. Data field words
4. Free-running timer (Releases internal lock —optional)
Upon reading the control and status word, if the BUSY bit is set in the CODE field, then the CPU should defer the access to the MB until this bit is negated. Reading the free running timer is not mandatory. If not executed the MB remains locked, unless the CPU reads the C/S word of another MB. Note that only a single MB is locked at a time. The only mandatory CPU read operation is the one on the control and status word to assure data coherency.
“Interrupt Flag Register (IFLAG)”
), and not by the control/status word CODE field for that MB. Polling the CODE field does not work because once a frame was received and the CPU services the MB (by reading the C/S word followed by unlocking the MB), the CODE field will not return to EMPTY. It will
remain FULL, as explained in Table 30-12
. If the CPU tries to workaround this behavior by writing to the
C/S word to force an EMPTY code after reading the MB, the MB is actually deactivated from any currently ongoing matching process. As a result, a newly received frame matching the ID of that MB may be lost.
In summary, never do polling by directly reading the C/S word of the MBs. Instead, read the IFLAG register.
30-22
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
FlexCAN
Note that the received identifier field is always stored in the matching MB, thus the contents of the ID field in an MB may change if the match was due to masking.
30.4.3.1
Self-Received Frames
Self-received frames are frames that are sent by the FlexCAN and received by itself. The FlexCAN sends a frame externally through the physical layer onto the CAN bus, and if the ID of the frame matches the ID of the FlexCAN MB, then the frame will be received by the FlexCAN. Such a frame is a self-received frame. Note that FlexCAN does not receive frames transmitted by itself if another device on the CAN bus has an ID that matches the FlexCAN Rx MB ID.
30.4.4
Matching Process
The matching process is an algorithm that scans the entire MB memory looking for Rx MBs programmed with the same ID as the one received from the CAN bus. Only MBs programmed to receive will participate in the matching process for received frames.
While the ID, DLC and data fields are retrieved from the CAN bus, they are stored temporarily in the serial message buffer (
Section 30.4.5.1, “Serial Message Buffers (SMBs)”
). The matching process takes place during the CRC field. If a matching ID is found in one of the MBs, the contents of the SMB will be transferred to the matched MB during the sixth bit of the end-of-frame field of the CAN protocol. This operation is called ‘move-in.’ If any protocol error (CRC, ACK, etc.) is detected, than the move-in operation does not happen.
An MB with a matching ID is free to receive a new frame if the MB is not locked (see
“Locking and Releasing Message Buffers”
). The CODE field is either EMPTY, FULL, or OVERRUN but the CPU has already serviced the MB (read the C/S word and then unlocked the MB).
Matching to a range of IDs is possible by using ID acceptance masks (RXGMASK, RX14MASK, and
RX15MASK). During the matching algorithm, if a mask bit is asserted, then the corresponding ID bit is compared. If the mask bit is negated, the corresponding ID bit is ‘don’t care.’
30.4.5
Message Buffer Handling
In order to maintain data coherency and FlexCAN proper operation, the CPU must obey the rules described
in Section 30.4.1, “Transmit Process”
and Section 30.4.3, “Receive Process.”
Any form of CPU accessing a MB structure within FlexCAN other than those specified may cause FlexCAN to behave in an unpredictable way.
30.4.5.1
Serial Message Buffers (SMBs)
To allow double buffering of messages, the FlexCAN has two shadow buffers called serial message buffers. These two buffers are used by the FlexCAN for buffering both received messages and messages to be transmitted. Only one SMB is active at a time, and its function depends upon the operation of the
FlexCAN at that time. At no time does the user have access to or visibility of these two buffers.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 30-23
FlexCAN
30.4.5.2
Message Buffer Deactivation
If the CPU wants to change the function of an active MB, the recommended procedure is to put the module into freeze mode and then change the CODE field of that MB. This is a safe procedure because the
FlexCAN waits for pending CAN bus and MB moving activities to finish before entering freeze mode.
Nevertheless, a mechanism is provided to maintain data coherence when the CPU writes to the control and status word of active MBs out of freeze mode.
Any CPU write access to the C/S word of an MB causes that MB to be excluded from the transmit or receive processes during the current matching or arbitration round. This mechanism is called MB deactivation. It is temporary, affecting only for the current match/arbitration round.
The purpose of deactivation is data coherency. The match/arbitration process scans the MBs to decide which MB to transmit or receive. If the CPU updates the MB in the middle of a match or arbitration process, the data of that MB may no longer be coherent; therefore, that MB is deactivated.
Even with the coherence mechanism described above, writing to the C/S word of active MBs when not in freeze mode may produce undesirable results. Examples are:
• Matching and arbitration are one-pass processes. If MBs are deactivated after they are scanned, no re-evaluation is done to determine a new match/winner. If an Rx MB with a matching ID is deactivated during the matching process after it was scanned, then this MB is marked as invalid to receive the frame, and FlexCAN will keep looking for another matching MB within the ones it has not scanned yet. If it can not find one, then the message will be lost. Suppose, for example, that two
MBs have a matching ID to a received frame, and the user deactivated the first matching MB after
FlexCAN has scanned the second. The received frame will be lost even if the second matching MB was free to receive.
• If a Tx MB containing the lowest ID is deactivated after the FlexCAN has scanned it, then the
FlexCAN will look for another winner within the MBs that it has not yet scanned. Therefore, it may transmit an MB that may not have the lowest ID at the time because a lower ID might be present that it had already scanned before the deactivation.
• There is a point in time until which the deactivation of a Tx MB causes it not to be transmitted (end of move-out). After this point, it is transmitted, but no interrupt is issued and the CODE field is not updated.
30.4.5.3
Locking and Releasing Message Buffers
Besides message buffer deactivation, the lock/release/busy mechanism is designed to guarantee data coherency during the receive process. The following examples demonstrate how the lock/release/busy mechanism will affect FlexCAN operation:
1. Reading a control/status word of a message buffer triggers a lock for that message buffer. A new received message frame that matches the message buffer cannot be written into this message buffer while it is locked.
2. To release a locked message buffer, the CPU either locks another message buffer (by reading its control/status word) or globally releases any locked message buffer (by reading the free-running timer).
30-24
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
FlexCAN
3. If a receive frame with a matching ID is received during the time the message buffer is locked, the receive frame will not be immediately transferred into that message buffer, but will remain in the
SMB. There is no indication when this occurs.
4. When a locked message buffer is released, if a frame with a matching identifier exists within the
SMB, then this frame will be transferred to the matching message buffer.
5. If two or more receive frames with matching IDs are received while a message buffer with a matching ID is locked, the last received frame with that ID is kept within the serial message buffer, while all preceding ones are lost. There is no indication of lost messages when this occurs.
6. If the user reads the control/status word of a receive message buffer while a frame is being transferred from a serial message buffer, the BUSY code will be indicated. The user should wait until this code is cleared before continuing to read from the message buffer to ensure data coherency. In this situation, the read of the control/status word will not lock the message buffer.
Polling the control/status word of a receive message buffer can lock it, preventing a message from being transferred into that buffer. If the control/status word of a receive message buffer is read, it should then be followed by a read of the control/status word of another buffer, or by reading the free-running timer, to ensure that the locked buffer is unlocked.
NOTE
Deactivation takes precedence over locking. If the CPU deactivates a locked
Rx MB, then its lock status is negated, and the MB is marked as invalid for the current matching round. Any pending message on the SMB will not be transferred to the MB anymore.
30.4.6
CAN Protocol Related Frames
30.4.6.1
Remote Frames
The remote frame is a message frame that is transmitted to request a data frame. The FlexCAN can be configured to transmit a data frame automatically in response to a remote frame, or to transmit a remote frame and then wait for the responding data frame to be received.
When transmitting a remote frame, the user initializes a message buffer as a transmit message buffer with the RTR bit set. Once this remote frame is transmitted successfully, the transmit message buffer automatically becomes a receive message buffer, with the same ID as the remote frame that was transmitted.
When a remote frame is received by the FlexCAN, the remote frame ID is compared to the IDs of all transmit message buffers programmed with a CODE of 1010. If there is an exact matching ID, the data frame in that message buffer is transmitted. If the RTR bit in the matching transmit message buffer is set, the FlexCAN will transmit a remote frame as a response.
A received remote frame is not stored in a receive message buffer. It is only used to trigger the automatic transmission of a frame in response. The mask registers are not used in remote frame ID matching. All ID bits (except RTR) of the incoming received frame must match for the remote frame to trigger a response transmission. The matching message buffer immediately enters the internal arbitration process, but is
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 30-25
FlexCAN considered as a normal Tx MB, with no higher priority. The data length of this frame is independent of the data length code (DLC) field in the remote frame that initiated its transmission.
30.4.6.2
Overload Frames
Overload frame transmissions are not initiated by the FlexCAN unless certain conditions are detected on the CAN bus. These conditions include detection of a dominant bit in the following:
• First or second bit of intermission
• Seventh (last) bit of the end-of-frame (EOF) field in receive frames
• Eighth (last) bit of the error frame delimiter or overload frame delimiter
30.4.7
Time Stamp
The value of TIMER is sampled at the beginning of the identifier field on the CAN bus. For a message being received, the time stamp will be stored in the TIMESTAMP entry of the receive message buffer at the time the message is written into that buffer. For a message being transmitted, the TIMESTAMP entry will be written into the transmit message buffer once the transmission has completed successfully.
The free-running timer can optionally be reset upon the reception of a frame into message buffer 0. This feature allows network time synchronization to be performed. See the CANCTRL[TSYN] bit.
30.4.8
Bit Timing
The FlexCAN module CANCTRL register configures the bit timing parameters required by the CAN protocol. The CLK_SRC, PRESDIV, RJW, PSEG1, PSEG2, and the PROPSEG fields allow the user to configure the bit timing parameters.
The CANCTRL[CLK_SRC] bit defines whether the module uses the internal bus clock or the output of the crystal oscillator via the EXTAL pin. The crystal oscillator clock should be selected whenever a tight tolerance (up to 0.1%) is required for the CAN bus timing. The crystal oscillator clock has better jitter performance than PLL generated clocks. The value of this bit should not be changed, unless the module is in disable mode (CANMCR[MDIS] bit is set)
The PRESDIV field controls a prescaler that generates the serial clock (S-clock), whose period defines the time quantum used to compose the CAN waveform. A time quantum is the atomic unit of time handled by the CAN engine.
CANCTRL[CLK_SRC]
Internal Bus Clock
(f sys/23
)
1
Prescaler
(1 .. 256)
Oscillator Clock (EXTAL)
0
Figure 30-14. CAN Engine Clocking Scheme f
Tq
=
( f sys/2
or EXTAL
PRESDIV + 1 )
S clock
Eqn. 30-3
30-26
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
FlexCAN f
Tq
=
A bit time is subdivided into three segments f
(
1
or EXTAL
PRESDIV + 1
(see
)
):
Eqn. 30-4
• SYNC_SEG: Has a fixed length of one time quantum. Signal edges are expected to happen within this section.
• Time Segment 1: Includes the propagation segment and the phase segment 1 of the CAN standard.
It can be programmed by setting the PROPSEG and the PSEG1 fields of the CANCTRL register so that their sum (plus 2) is in the range of 4 to 16 time quanta.
• Time Segment 2: Represents the phase segment 2 of the CAN standard. It can be programmed by setting the PSEG2 field of the CANCTRL register (plus 1) to be 2 to 8 time quanta long.
Bit Rate = f
(number of Time Quanta)
Eqn. 30-5
NRZ Signal
SYNC_SEG
1
Time Segment 1
(PROP_SEG + PSEG1 + 2)
4 ... 16
8 ... 25 Time Quanta
= 1 Bit Time
Time Segment 2
(PSEG2 + 1)
2 ... 8
Transmit Point Sample Point
(single or triple sampling)
Figure 30-15. Segments within the Bit Time
Table 30-14. Time Segment Syntax
Syntax Description
SYNC_SEG
Transmit Point
Sample Point
System expects transitions to occur on the bus during this period.
A node in transmit mode transfers a new value to the CAN bus at this point.
A node samples the bus at this point. If the three samples per bit option is selected, then this point marks the position of the third sample.
gives an overview of the CAN compliant segment settings and the related parameter values.
1. For further explanation of the underlying concepts please refer to ISO/DIS 11519 – 1, Section 10.3. Reference also the Bosch
CAN 2.0A/B protocol specification dated September 1991 for bit timing.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 30-27
FlexCAN
NOTE
It is the user’s responsibility to ensure the bit time settings are in compliance with the CAN standard. For bit time calculations, use an IPT (Information
Processing Time) of 2, which is the value implemented in the FlexCAN module
Table 30-15. CAN Standard Compliant Bit Time Segment Settings
Time Segment 1
5 .. 10
4 .. 11
5 .. 12
6 .. 13
7 .. 14
8 .. 15
9 .. 16
Time Segment 2
6
7
8
4
5
2
3
Re-synchronization
Jump Width
1 .. 2
1 .. 3
1 .. 4
1 .. 4
1 .. 4
1 .. 4
1 .. 4
30.5
FlexCAN Initialization Sequence
Initialization of the FlexCAN includes the initial configuration of the message buffers and configuration of the CAN communication parameters following a reset, as well as any reconfiguration that may be required during operation. The FlexCAN module may be reset in three ways:
• Device level hard reset—resets all memory mapped registers asynchronously
• Device level soft reset—resets some of the memory mapped registers synchronously (refer to
to see which registers are affected by soft reset)
• CANMCR[SOFT_RST] bit—has the same effect as the device level soft reset
Soft reset is synchronous and has to follow an internal request/acknowledge procedure across clock domains. Therefore, it may take some time to fully propagate its effects. The CANMCR[SOFT_RST] bit remains asserted while soft reset is pending, so software can poll this bit to know when the reset has completed. Also, soft reset can not be applied while clocks are shut down in any of the low power modes.
The low power mode should be exited and the clocks resumed before applying soft reset.
The clock source, CANCTRL[CLK_SRC], should be selected while the module is in disable mode. After the clock source is selected and the module is enabled (CANMCR[MDIS] bit cleared), the FlexCAN automatically enters freeze mode. In freeze mode, the FlexCAN is un-synchronized to the CAN bus, the
CANMCR register’s HALT and FRZ bits are set, the internal state machines are disabled, and the
CANMCR register’s FRZ_ACK and NOT_RDY bits are set. The CANTX pin is in recessive state and the
FlexCAN does not initiate any transmission or reception of CAN frames. Note that the message buffers are not affected by reset, so they are not automatically initialized.
30-28
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
FlexCAN
). The following is a generic initialization sequence applicable to the FlexCAN module:
1. Initialize all operation modes in the CANCTRL register.
a) Initialize the bit timing parameters PROPSEG, PSEGS1, PSEG2, and RJW.
b) Select the S-clock rate by programming the PRESDIV field.
c) Select the internal arbitration mode via the LBUF bit.
2. Initialize message buffers.
a) The control/status word of all message buffers must be written as either an active or inactive message buffer.
b) All other entries in each message buffer should be initialized as required.
3. Initialize RXGMASK, RX14MASK, and RX15MASK registers for acceptance mask as needed.
4. Initialize FlexCAN interrupt handler.
a) Initialize the interrupt controller registers for any needed interrupts. See Chapter 14, “Interrupt
Controller Modules,” for more information.
b) Set the required mask bits in the IMASK register (for all message buffer interrupts) and the
CANCTRL (for bus off and error interrupts).
5. Clear the CANMCR[HALT] bit. At this point, the FlexCAN will attempt to synchronize with the
CAN bus.
30.5.1
Interrupts
There are three interrupt sources for the FlexCAN module. A combined interrupt for all 16 MBs is generated by combining all the interrupt sources from MBs. This interrupt gets generated when any of the
16 MB interrupt sources generates a interrupt. In this case, the CPU must read the IFLAG register to determine which MB caused the interrupt. The other two interrupt sources (bus off and error) act in the same way, and are located in the ERRSTAT register. The bus off and error interrupt mask bits are located in the CANCTRL register.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor Preliminary 30-29
FlexCAN
30-30
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Preliminary Freescale Semiconductor
Chapter 31
Debug Module
31.1
Introduction
This chapter describes the Revision B+ enhanced hardware debug module.
31.1.1
Overview
The debug module is shown in
.
ColdFire CPU Core
High-speed core bus
(f sys
)
Debug Module
Control
BKPT
Trace Port
PST[3:0], DDATA[3:0]
PSTCLK
Communication Port
DSCLK, DSI, DSO
Figure 31-1. Processor/Debug Module Interface
Depending upon the package, some devices contain only the ALLPST signal and do not have the PST[3:0] and DDATA[3:0] signals. ALLPST is a logical ‘AND’ of the PST[3:0] signals, and when asserted reflects that the core is halted.
Debug support is divided into three areas:
• Real-time trace support—The ability to determine the dynamic execution path through an application is fundamental for debugging. The ColdFire solution implements an 8-bit parallel output bus that reports processor execution status and data to an external emulator system. See
Section 31.3, “Real-Time Trace Support .”
• Background debug mode (BDM)—Provides low-level debugging in the ColdFire processor complex. In BDM, the processor complex is halted and a variety of commands can be sent to the processor to access memory, registers, and peripherals. The external emulator uses a three-pin, serial, full-duplex communication port. See
Section 31.5, “Background Debug Mode (BDM)
,” and
Section 31.4, “Memory Map/Register Definition .”
• Real-time debug support—BDM requires the processor to be halted, which many real-time embedded applications cannot do. External development systems can access memory because the hardware supports concurrent operation of the processor and BDM-initiated commands. Debug interrupts let real-time systems execute a unique service routine that can quickly save the contents
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-1
Debug Module of key registers and variables and return the system to normal operation. See
.
31.1.1.1
The New Debug Module Hardware (Rev. B+)
The revision B+ debug module features a small enhancement over revision B: the addition of three PC breakpoint registers (PCBR1–3). These new registers are mapped to DRc[3:0] addresses 0x18, 0x1A, and
0x1B. Additional PC breakpoints enable software debugging. However, there are no masking registers associated with these new registers.
31.1.1.2
Enhancements over Revision A
The new debug hardware also contains the same enhancements that revision B has over revision A, while maintaining backwards compatibility with revision A. These enhancements are discussed below.
The revision B/B+ implementation has added registers that eliminate restrictions between BDM commands and the use of the hardware breakpoint logic. In some cases, the additional hardware is not program-visible; in other cases, there have been extensions to the debug module programming model.
The register containing the BDM memory address is not a program-visible resource. Rather, it is a register loaded automatically during the execution of a BDM command. In the Rev. B design, the execution of a
BDM command does not affect the hardware breakpoint logic unless those registers are specifically accessed.
The other register added to the debug module programming model is the BDM address attribute register
(BAAR). The BAAR is mapped to a DRc[3:0] address of 0x05. This 8-bit register is equivalent in format
of the low-order byte of the AATR register (See Section 31.4.3, “BDM Address Attribute (BAAR)”
). This register specifies the memory space attributes associated with all BDM memory-referencing commands.
Additionally, a bit was added to the CSR register (CSR[BKD]) that configures the debug module to assert or not assert an interrupt to the processor when the BKPT signal is asserted. See
“Configuration/Status Register (CSR).”
The level 1 and level 2 triggers are also configurable to trigger on either an AND or an OR condition. The revision A debug module only triggers on an AND condition.
31.2
E
xternal Signal Description
describes debug module signals. All ColdFire debug signals are unidirectional and related to a rising edge of the processor’s clock signal. The standard 26-pin debug connector is shown in
“Freescale-Recommended BDM Pinout .”
31-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
Table 31-1. Debug Module Signals
Signal Description
Development Serial
Clock (DSCLK)
Internally synchronized input. (The logic level on DSCLK is validated if it has the same value on two consecutive rising PSTCLK edges.) Clocks the serial communication port to the debug module during packet transfers. Maximum frequency is 1/5 the processor status clock (PSTCLK) speed. At the synchronized rising edge of DSCLK, the data input on DSI is sampled and DSO changes state.
Development Serial
Input (DSI)
Internally synchronized input that provides data input for the serial communication port to the debug module.
Development Serial
Output (DSO)
Provides serial output communication for debug module responses. DSO is registered internally.
Breakpoint (BKPT) Input used to request a manual breakpoint. Assertion of BKPT puts the processor into a halted state after the current instruction completes. Halt status is reflected on processor status signals
(PST[3:0]) as the value 0xF. Also, in Rev B and B+ if the CSR[BKD] bit is set, the assertion of the
BKPT signal will generate a debug interrupt exception to the processor.
Debug Data
(DDATA[3:0])
These output signals display the register breakpoint status as a default, or optionally, captured address and operand values. The capturing of data values is controlled by the setting of the CSR.
Additionally, execution of the WDDATA instruction by the processor captures operands which are displayed on DDATA. These signals are updated each processor cycle. These signals are not implemented on packages containing fewer than 100 pins.
Processor Status
(PST[3:0])
All Processor
Status Outputs
(ALLPST)
These output signals report the processor status.
Table 31-2 shows the encoding of these
signals. These outputs indicate the current status of the processor pipeline and, as a result, are not related to the current bus transfer. The PST value is updated each processor cycle. These signals are not implemented on packages containing fewer than 100 pins.
ALLPST is a logical ‘AND’ of the four PST signals is provided on all packages. PST[3:0] and
DDATA[3:0] are not available on the low cost (less than 100 pin) packages. When asserted, reflects that the core is halted.
shows PSTCLK timing with respect to PST and DDATA.
PST or DDATA
Figure 31-2. PSTCLK Timing
31.3
Real-Time Trace Support
Real-time trace, which defines the dynamic execution path, is a fundamental debug function. The ColdFire solution is to include a parallel output port providing encoded processor status and data to an external development system. This port is partitioned into two 4-bit nibbles: one nibble allows the processor to transmit processor status, (PST), and the other allows operand data to be displayed (debug data , DDATA).
The processor status may not be related to the current bus transfer.
External development systems can use PST outputs with an external image of the program to completely track the dynamic execution path. This tracking is complicated by any change in flow, where branch target
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-3
Debug Module address calculation is based on the contents of a program-visible register (variant addressing). DDATA outputs can be configured to display the target address of such instructions in sequential nibble increments across multiple processor clock cycles, as described in
Section 31.3.1, “Begin Execution of Taken Branch
.” Two 32-bit storage elements form a FIFO buffer connecting the processor’s high-speed local bus to the external development system through PST[3:0] and DDATA[3:0]. The buffer captures branch target addresses and certain data values for eventual display on the DDATA port, one nibble at a time starting with the least significant bit (lsb).
Execution speeds affected only when both storage elements contain valid data to be dumped to the DDATA port. The core stalls until one FIFO entry is available.
shows the encoding of these signals.
Table 31-2. Processor Status Encoding
PST[3:0]
Definition
Hex Binary
0x0 0000 Continue execution. Many instructions execute in one processor cycle. If an instruction requires more processor clock cycles, subsequent clock cycles are indicated by driving PST outputs with this encoding.
0x1 0001 Begin execution of one instruction. For most instructions, this encoding signals the first processor clock cycle of an instruction’s execution. Certain change-of-flow opcodes, plus the PULSE and WDDATA instructions, generate different encodings.
0x2 0010 Reserved
0x3 0011 Entry into user-mode. Signaled after execution of the instruction that caused the ColdFire processor to enter user mode.
0x4 0100 Begin execution of PULSE and WDDATA instructions. PULSE defines logic analyzer triggers for debug and/or performance analysis. WDDATA lets the core write any operand (byte, word, or longword) directly to the DDATA port, independent of debug module configuration. When WDDATA is executed, a value of 0x4 is signaled on the PST port, followed by the appropriate marker, and then the data transfer on the DDATA port. Transfer length depends on the WDDATA operand size.
0x5 0101 Begin execution of taken branch. For some opcodes, a branch target address may be displayed on
DDATA depending on the CSR settings. CSR also controls the number of address bytes displayed, indicated by the PST marker value preceding the DDATA nibble that begins the data output. See
Section 31.3.1, “Begin Execution of Taken Branch (PST = 0x5).”
Also indicates that the SYNC_PC command has been issued.
0x6 0110 Reserved
0x7 0111 Begin execution of return from exception (RTE) instruction.
0x8–
0xB
1000–
1011
Indicates the number of bytes to be displayed on the DDATA port on subsequent processor clock cycles. The value is driven onto the PST port one PSTCLK cycle before the data is displayed on
DDATA.
0x8 Begin 1-byte transfer on DDATA.
0x9 Begin 2-byte transfer on DDATA.
0xA Begin 3-byte transfer on DDATA.
0xB Begin 4-byte transfer on DDATA.
0xC 1100 Exception processing. Exceptions that enter emulation mode (debug interrupt or optionally trace) generate a different encoding, as described below. Because the 0xC encoding defines a multiple-cycle mode, PST outputs are driven with 0xC until exception processing completes.
31-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
Table 31-2. Processor Status Encoding (continued)
PST[3:0]
Definition
Hex Binary
0xD 1101 Entry into emulator mode.
Displayed during emulation mode (debug interrupt or optionally trace).
Because this encoding defines a multiple-cycle mode, PST outputs are driven with 0xD until exception processing completes.
0xE 1110 Processor is stopped. Appears in multiple-cycle format when the processor executes a STOP instruction. The ColdFire processor remains stopped until an interrupt occurs, thus PST outputs display 0xE until the stopped mode is exited
.
0xF 1111 Processor is halted. Because this encoding defines a multiple-cycle mode, the PST outputs display
0xF until the processor is restarted or reset. (see Section 31.5.1, “CPU Halt ”)
31.3.1
Begin Execution of Taken Branch (PST = 0x5)
PST is 0x5 when a taken branch is executed. For some opcodes, a branch target address may be displayed on DDATA depending on the CSR settings. CSR also controls the number of address bytes displayed, which is indicated by the PST marker value immediately preceding the DDATA nibble that begins the data output.
Bytes are displayed in least-to-most-significant order. The processor captures only those target addresses associated with taken branches which use a variant addressing mode; that is, RTE and RTS instructions,
JMP and JSR instructions using address register indirect or indexed addressing modes, and all exception vectors.
The simplest example of a branch instruction using a variant address is the compiled code for a C language case statement. Typically, the evaluation of this statement uses the variable of an expression as an index into a table of offsets, where each offset points to a unique case within the structure. For such change-of-flow operations, the ColdFire processor uses the debug pins to output the following sequence of information on successive processor clock cycles:
1. Use PST (0x5) to identify that a taken branch was executed.
2. Using the PST pins, optionally signal the target address to be displayed sequentially on the DDATA pins. Encodings 0x9–0xB identify the number of bytes displayed .
3. The new target address is optionally available on subsequent cycles using the DDATA port. The number of bytes of the target address displayed on this port is configurable (2, 3, or 4 bytes).
Another example of a variant branch instruction would be a JMP (A0) instruction.
shows the
PST and DDATA outputs that indicate a JMP (A0) execution (assuming the CSR was programmed to display the lower 2 bytes of an address).
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-5
Debug Module
PSTCLK
PST 0x5 0x9 default default default default
DDATA 0x0 0x0 A[3:0] A[7:4] A[11:8] A[15:12]
Figure 31-3. Example JMP Instruction Output on PST/DDATA
PST 0x5 indicates a taken branch and the marker value 0x9 indicates a 2-byte address. Thus, the subsequent 4 nibbles of DDATA display the lower 2 bytes of address register A0 in least-to-most-significant nibble order. The PST output after the JMP instruction completes depends on the target instruction. The PST can continue with the next instruction before the address has completely displayed on DDATA because of the DDATA FIFO. If the FIFO is full and the next instruction has captured values to display on DDATA, the pipeline stalls (PST = 0x0) until space is available in the FIFO.
31.4
Memory Map/Register Definition
In addition to the existing BDM commands that provide access to the processor’s registers and the memory subsystem, the debug module contains 19 registers to support the required functionality. These registers are also accessible from the processor’s supervisor programming model by executing the WDEBUG instruction (write only). Thus, the breakpoint hardware in the debug module can be written by the external development system using the debug serial interface or by the operating system running on the processor core. Software is responsible for guaranteeing that accesses to these resources are serialized and logically consistent. Hardware provides a locking mechanism in the CSR to allow the external development system to disable any attempted writes by the processor to the breakpoint registers (setting CSR[IPW]). BDM commands must not be issued if the ColdFire processor is using the WDEBUG instruction to access debug module registers, or the resulting behavior is undefined, while DSCLK is quiescent.
These registers, shown in Table 31-3 , are treated as 32-bit quantities, regardless of the number of
implemented bits. These registers are also accessed through the BDM port by the commands, WDMREG and RDMREG , described in
Section 31.5.3.3, “Command Set Descriptions .” These commands contain a
5-bit field, DRc, that specifies the register, as shown in
Table 31-3. Debug Module Memory Map
DRc[4–0]
0x00
0x05
0x06
0x07
0x08
0x09
Register
Configuration/Status Register (CSR)
BDM Address Attribute Register (BAAR)
Address Attribute Trigger Register (AATR)
Trigger Definition Register (TDR)
PC Breakpoint Register 0 (PBR0)
PC Breakpoint Mask Register (PBMR)
Width
(bits)
32
32
1
32
32
32
Access Reset Value Section/Page
See Note 0x0090_0000
See Note 0x05
See Note 0x0005
See Note 0x0000_0000
See Note Undefined
See Note Undefined
31-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
Table 31-3. Debug Module Memory Map (continued)
1
DRc[4–0] Register
Width
(bits)
Access Reset Value
0x0C
0x0D
0x0E
0x0F
0x18
0x1A
0x1B
Address High Breakpoint Register (ABHR)
Address Low Breakpoint Register (ABLR)
Data Breakpoint Register (DBR)
Data Breakpoint Mask Register (DBMR)
PC Breakpoint Register 1 (PBR1)
PC Breakpoint Register 2 (PBR2)
PC Breakpoint Register 3 (PBR3)
32
32
32
32
32
32
32
See Note
See Note
See Note
See Note
See Note
See Note
See Note
Undefined
Undefined
Undefined
Undefined
Undefined
Undefined
Undefined
Each debug register is accessed as a 32-bit register; reserved fields are not used (don’t care).
Section/Page
NOTE
Debug control registers can be written by the external development system or the CPU through the WDEBUG instruction. CSR is write-only from the programming model. It can be read or written through the BDM port using the
RDMREG
and
WDMREG
commands.
31.4.1
Shared Debug Resources
The debug module implementation provides a common hardware structure for both BDM and breakpoint functionality. Certain hardware structures are used for both BDM and breakpoint purposes as shown in
.
Table 31-4. Shared BDM/Breakpoint Hardware
Register
AATR
ABHR
DBR
BDM Function Breakpoint Function
Bus attributes for all memory commands Attributes for address breakpoint
Address for all memory commands Address for address breakpoint
Data for all BDM write commands Data for data breakpoint
Thus, loading a register to perform a specific function that shares hardware resources is destructive to the shared function. For example, if an operand address breakpoint is loaded into the debug module, a BDM command to access memory overwrites an address breakpoint in ABHR. If a data breakpoint is configured, a BDM write command overwrites the data breakpoint in DBR.
31.4.2
Configuration/Status Register (CSR)
The CSR defines the debug configuration for the processor and memory subsystem and contains status information from the breakpoint logic. CSR is write-only from the programming model. It can be read from and written to through the BDM port. CSR is accessible in supervisor mode as debug control register
0x00 using the WDEBUG instruction and through the BDM port using the RDMREG and WDMREG commands.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-7
Debug Module
DRc[4:0]: 0x00 (CSR)
31
R
W
Reset 0
30 29
BSTAT
0 0
28 27 26 25 24
FOF TRG HALT BKPT
23
0 0 0 0 0 1
22
HRL
21
0 0
20
Access: Supervisor write-only
BDM read/write
19
0
18 17 16
BKD PCD IPW
1 0 0 0
15 14 13
R
W
MAP TRC EMU
Reset 0 0 0
12
DDC
11 10
UHE
9 8 7
0
6 5 4
NPL IPI SSM BTB
0 0 0 0 0 0 0 0
Figure 31-4. Configuration/Status Register (CSR)
0
Table 31-5. CSR Field Descriptions
3
0
0
2
0
0
1
0
0
0
0
0
Field Description
27
FOF
26
TRG
25
HALT
24
BKPT
23–20
HRL
31–28
BSTAT
Breakpoint status. Provides read-only status (from the BDM port only) information concerning hardware breakpoints.
BSTAT is cleared by a TDR write or by a CSR read when either a level-2 breakpoint is triggered or a level-1 breakpoint is triggered and the level-2 breakpoint is disabled.
0000 No breakpoints enabled
0001 Waiting for level-1 breakpoint
0010 Level-1 breakpoint triggered
0101 Waiting for level-2 breakpoint
0110 Level-2 breakpoint triggered
Fault-on-fault. If FOF is set, a catastrophic halt occurred and forced entry into BDM. FOF is cleared whenever CSR is read (from the BDM port only).
Hardware breakpoint trigger. If TRG is set, a hardware breakpoint halted the processor core and forced entry into
BDM. Reset, the debug GO command, or reading CSR (from the BDM port only) will clear TRG.
Processor halt. If HALT is set, the processor executed a HALT and forced entry into BDM. Reset, the debug GO command, or reading CSR (from the BDM port only) will clear HALT.
Breakpoint assert. If BKPT is set, BKPT was asserted, forcing the processor into BDM. Reset, the debug GO command, or reading CSR (from the BDM port only) will clear BKPT.
Hardware revision level. Indicates, from the BDM port only, the level of debug module functionality. An emulator could use this information to identify the level of functionality supported.
1001 Revision B+ (This is the only valid value for this processor)
19–18 Reserved, should be cleared.
17
PCD
PST/DDATA Disable. Disables the PST/DDATA output signal. PSTCLK is unaffected, it remains under the control of the DISCLK bit in the SYNCR register.
0 Normal operation
1 Disables the generation of the PSTDDATA output signals, and forces these signals to remain quiescent
16
IPW
15
MAP
Inhibit processor writes. Setting IPW inhibits processor-initiated writes to the debug module’s programming model registers. IPW can be modified only by commands from the external development system.
Force processor references in emulator mode.
0 All emulator-mode references are mapped into supervisor code and data spaces.
1 The processor maps all references while in emulator mode to a special address space, TT = 10, TM = 101 or 110.
31-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
Table 31-5. CSR Field Descriptions (continued)
Field
14
TRC
13
EMU
12–11
DDC
10
UHE
9–8
BTB
7
6
NPL
5
IPI
4
SSM
3–0
Description
Force emulation mode on trace exception. If TRC = 1, the processor enters emulator mode when a trace exception occurs. If TRC=0, the processor enters supervisor mode.
.”
Debug data control. Controls operand data capture for DDATA, which displays the number of bytes defined by the operand reference size before the actual data; byte displays 8 bits, word displays 16 bits, and long displays 32 bits
(one nibble at a time across multiple PSTCLK cycles). See
.
00 No operand data is displayed.
01 Capture all write data.
10 Capture all read data.
11 Capture all read and write data.
User halt enable. Selects the CPU privilege level required to execute the HALT instruction.
0 HALT is a supervisor-only instruction.
1 HALT is a supervisor/user instruction.
Branch target bytes. Defines the number of bytes of branch target address DDATA displays.
00 0 bytes
01 Lower 2 bytes of the target address
10 Lower 3 bytes of the target address
11 Entire 4-byte target address
See
Section 31.3.1, “Begin Execution of Taken Branch (PST = 0x5) .”
Reserved, should be cleared.
Non-pipelined mode. Determines whether the core operates in pipelined or mode or not.
0 Pipelined mode
1 Nonpipelined mode. The processor effectively executes one instruction at a time with no overlap. This adds at least 5 cycles to the execution time of each instruction. Given an average execution latency of 1.6 cycles/instruction, throughput in non-pipeline mode would be 6.6 cycles/instruction, approximately 25% or less of pipelined performance.
Regardless of the NPL state, a triggered PC breakpoint is always reported before the triggering instruction executes.
In normal pipeline operation, the occurrence of an address and/or data breakpoint trigger is imprecise. In non-pipeline mode, triggers are always reported before the next instruction begins execution and trigger reporting can be considered precise.
Ignore pending interrupts.
1 Core ignores any pending interrupt requests signalled while in single-instruction-step mode.
0 Core services any pending interrupt requests that were signalled while in single-step mode.
Single-step mode. Setting SSM puts the processor in single-step mode.
0 Normal mode.
1 Single-step mode. The processor halts after execution of each instruction. While halted, any BDM command can be executed. On receipt of the GO command, the processor executes the next instruction and halts again. This process continues until SSM is cleared.
Reserved, should be cleared.
31.4.3
BDM Address Attribute (BAAR)
The BAAR register defines the address space for memory-referencing BDM commands. BAAR[R, SZ] are loaded directly from the BDM command, while the low-order 5 bits can be programmed from the external development system. To maintain compatibility with the Rev. A implementation, this register is
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-9
Debug Module loaded any time the AATR is written. The BAAR is initialized to a value of 0x05, setting supervisor data as the default address space.
DRc[4:0]: 0x05 (BAAR)
7 6 5 4 3 2
Access: Supervisor write-only
BDM write-only
1 0
R
W
Reset:
R
0
SZ TT
0 0 0 0 1
Figure 31-5. BDM Address Attribute Register (BAAR)
Table 31-6. BAAR Field Description
TM
0 1
Field
7
R
6–5
SZ
4–3
TT
2–0
TM
Description
Read/Write.
0 Write
1 Read
Size.
00 Longword
01 Byte
10 Word
11 Reserved
Transfer type. See the TT definition in the AATR description,
Section 31.4.4, “Address Attribute Trigger Register
Transfer modifier. See the TM definition in the AATR description,
Section 31.4.4, “Address Attribute Trigger Register
31.4.4
Address Attribute Trigger Register (AATR)
The AATR defines address attributes and a mask to be matched in the trigger. The register value is compared with address attribute signals from the processor’s local high-speed bus, as defined by the setting of the trigger definition register (TDR). AATR is accessible in supervisor mode as debug control register 0x06 using the WDEBUG instruction and through the BDM port using the WDMREG command.
DRc[4:0]: 0x06 (AATR)
14 13 12 11 10 9 8 7 6 5 4
Access: Supervisor write-only
BDM write-only
3 2 1 0 15
R
W RM
Reset 0 0
SZM
0 0
TTM
0 0
TMM
0 0
R
0 0
SZ
0 0
Figure 31-6. Address Attribute Trigger Register (AATR)
TT
0 1
TM
0 1
31-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
Table 31-7. AATR Field Descriptions
Field
15
RM
14–13
SZM
12–11
TTM
10–8
TMM
7
R
6–5
SZ
4–3
TT
2–0
TM
Description
Read/write mask. Setting RM masks R in address comparisons.
Size mask. Setting an SZM bit masks the corresponding SZ bit in address comparisons.
Transfer type mask. Setting a TTM bit masks the corresponding TT bit in address comparisons.
Transfer modifier mask. Setting a TMM bit masks the corresponding TM bit in address comparisons.
Read/write. R is compared with the R/W signal of the processor’s local bus.
Size. Compared to the processor’s local bus size signals.
00 Longword
01 Byte
10 Word
11 Reserved
Transfer type. Compared with the local bus transfer type signals.
00 Normal processor access
01 Reserved
10 Emulator mode access
11 Acknowledge/CPU space access
These bits also define the TT encoding for BDM memory commands. In this case, the 01 encoding generates an external master or DMA access (for backward compatibility). These bits are used to decode the TM bits.
Transfer modifier. Compared with the local bus transfer modifier signals, which give supplemental information for each transfer type. These bits also define the TM encoding for BDM memory commands (for backward compatibility).
TM
000
001
010
011
100
101
110
111
TT=00
(normal mode)
Reserved
User data access
User code access
Reserved
Reserved
Supervisor data access
Supervisor code access
Reserved
TT=10
(emulator mode)
Reserved
Reserved
Reserved
Reserved
Reserved
Emulator mode access
Emulator code
Access
Reserved
TT=11
(acknowledge/CPU space transfers)
CPU space access
Interrupt ack level 1
Interrupt ack level 2
Interrupt ack level 3
Interrupt ack level 4
Interrupt ack level 5
Interrupt ack level 6
Interrupt ack level 7
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-11
Debug Module
31.4.5
Trigger Definition Register (TDR)
The TDR configures the operation of the hardware breakpoint logic that corresponds with the
ABHR/ABLR/AATR, PBR/PBMR, and DBR/DBMR registers within the debug module. The TDR controls the actions taken under the defined conditions. Breakpoint logic may be configured as a one- or two-level trigger. TDR[31–16] bits define the second-level trigger and TDR[15–0] bits define the first-level trigger.
NOTE
The debug module has no hardware interlocks, so to prevent spurious breakpoint triggers while the breakpoint registers are being loaded, disable
TDR (by clearing TDR[29,13]) before defining triggers.
A write to TDR clears the CSR trigger status bits, CSR[BSTAT]. TDR is accessible in supervisor mode as debug control register 0x07 using the WDEBUG instruction and through the BDM port using the WDMREG command.
DRc[4:0]: 0x07 (TDR) Access: Supervisor write-only
BDM write-only
28 27 26 25 24
Second Level Trigger
23 22 21 20 19 18 17 16 31 30
R
W
Reset 0
TRC
0
29
L2EBL
0 0
12
0
11
0
10
L2ED
0
9
0 0 0
First Level Trigger
L2DI
0
8 7 6 5
0
4
L2EA
0
3
0
2
L2EPC L2PCI
0 0
1 0 15 14 13
R
W L2T L1T L1EBL
Reset 0 0 0 0 0 0
L1ED
0 0 0 0
L1DI
0
Figure 31-7. Trigger Definition Register (TDR)
0
L1EA
0 0
L1EPC L1PCI
0 0
Table 31-8. TDR Field Descriptions
Field Description
31–30
TRC
29
L2EBL
Trigger response control. Determines how the processor responds to a completed trigger condition. The trigger response is always displayed on DDATA.
00 Display on DDATA only
01 Processor halt
10 Debug interrupt
11 Reserved
Enable level 2 breakpoint. Global enable for the breakpoint trigger.
0 Disables all level 2 breakpoints
1 Enables all level 2 breakpoint triggers
31-12
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
Table 31-8. TDR Field Descriptions (continued)
Field
28–22
L2ED
Description
Enable level 2 data breakpoint. Setting an L2ED bit enables the corresponding data breakpoint condition based on the size and placement on the processor’s local data bus. Clearing all ED bits disables data breakpoints.
TDR Bit
24
23
22
28
27
26
25
Description
Data longword. Entire processor’s local data bus.
Lower data word.
Upper data word.
Lower lower data byte. Low-order byte of the low-order word.
Lower middle data byte. High-order byte of the low-order word.
Upper middle data byte. Low-order byte of the high-order word.
Upper upper data byte. High-order byte of the high-order word.
21
L2DI
20–18
L2EA
Level 2 data breakpoint invert. Inverts the logical sense of all the data breakpoint comparators. This can develop a trigger based on the occurrence of a data value other than the DBR contents.
0 No inversion
1 Invert data breakpoint comparators.
Enable level 2 address breakpoint. Setting an L2EA bit enables the corresponding address breakpoint. Clearing all three bits disables the breakpoint.
TDR Bit
20
19
18
Description
Address breakpoint inverted. Breakpoint is based outside the range between ABLR and ABHR.
Address breakpoint range. The breakpoint is based on the inclusive range defined by ABLR and ABHR.
Address breakpoint low. The breakpoint is based on the address in the ABLR.
17
L2EPC
Enable level 2 PC breakpoint.
0 Disable PC breakpoint
1 Enable PC breakpoint
16
L2PCI
15
L2T
Level 2 PC breakpoint invert.
0 The PC breakpoint is defined within the region defined by PBR and PBMR.
1 The PC breakpoint is defined outside the region defined by PBR and PBMR.
Level 2 trigger. Determines the logic operation for the trigger between the PC_condition and the (Address_range &
Data_condition) where the inclusion of a Data_condition is optional. The ColdFire debug architecture supports the creation of single or double-level triggers.
0 Level 2 trigger = PC_condition & Address_range & Data_condition
1 Level 2 trigger = PC_condition | (Address_range & Data_condition)
Note: Debug Rev A only had the ‘AND’ condition available for the triggers.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-13
Debug Module
Table 31-8. TDR Field Descriptions (continued)
Field Description
14
L1T
13
L1EBL
12–6
L1ED
Level 1 trigger. Determines the logic operation for the trigger between the PC_condition and the (Address_range &
Data_condition) where the inclusion of a Data_condition is optional. The ColdFire debug architecture supports the creation of single or double-level triggers.
0 Level 1 trigger = PC_condition & Address_range & Data_condition
1 Level 1 trigger = PC_condition | (Address_range & Data_condition)
Note: Debug Rev A only had the ‘AND’ condition available for the triggers.
Enable level 1 breakpoint. Global enable for the breakpoint trigger.
0 Disables all level 1 breakpoints
1 Enables all level 1 breakpoint triggers
Enable level 1 data breakpoint. Setting an L1ED bit enables the corresponding data breakpoint condition based on the size and placement on the processor’s local data bus. Clearing all L1ED bits disables data breakpoints.
TDR Bit
8
7
6
12
11
10
9
Description
Data longword. Entire processor’s local data bus.
Lower data word.
Upper data word.
Lower lower data byte. Low-order byte of the low-order word.
Lower middle data byte. High-order byte of the low-order word.
Upper middle data byte. Low-order byte of the high-order word.
Upper upper data byte. High-order byte of the high-order word.
5
L1DI
4–2
L1EA
Level 1 data breakpoint invert. Inverts the logical sense of all the data breakpoint comparators. This can develop a trigger based on the occurrence of a data value other than the DBR contents.
0 No inversion
1 Invert data breakpoint comparators.
Enable level 1 address breakpoint. Setting an L1EA bit enables the corresponding address breakpoint. Clearing all three bits disables the address breakpoint.
TDR Bit
4
3
2
Description
Enable address breakpoint inverted. Breakpoint is based outside the range between ABLR and ABHR.
Enable address breakpoint range. The breakpoint is based on the inclusive range defined by ABLR and ABHR.
Enable address breakpoint low. The breakpoint is based on the address in the ABLR.
1
L1EPC
Enable level 1 PC breakpoint.
0 Disable PC breakpoint
1 Enable PC breakpoint
0
L1PCI
Level 1 PC breakpoint invert.
0 The PC breakpoint is defined within the region defined by PBR and PBMR.
1 The PC breakpoint is defined outside the region defined by PBR and PBMR.
31-14
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
31.4.6
Program Counter Breakpoint/Mask Registers (PBR, PBMR)
The PBR register defines an instruction address for use as part of the trigger. This register’s contents are compared with the processor’s program counter register when TDR is configured appropriately. PBR bits are masked by setting corresponding PBMR bits. Results are compared with the processor’s program
counter register, as defined in TDR. Figure 31-8 shows the PC breakpoint register.
The PC breakpoint register is accessible in supervisor mode using the WDEBUG instruction and through the BDM port using the RDMREG and WDMREG
commands and values shown in Section 31.5.3.3,
DRc[4:0]: 0x08 (PBR) Access: Supervisor write-only
BDM read/write
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
R
W Address
Reset – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –
Figure 31-8. Program Counter Breakpoint Register (PBR)
Table 31-9. PBR Field Descriptions
Field Description
31–1
Address
PC breakpoint address. The address to be compared with the PC as a breakpoint trigger.
shows PBMR. PBMR is accessible in supervisor mode as debug control register 0x09 using the WDEBUG instruction and via the BDM port using the WDMREG command. PBMR only masks PBR0.
DRc[4:0]: 0x09 (PBMR) Access: Supervisor write-only
BDM write-only
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
R
W Mask
Reset – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –
Figure 31-9. Program Counter Breakpoint Mask Register (PBMR)
Field
31–0
Mask
Table 31-10. PBMR Field Descriptions
Description
PC breakpoint mask. A zero in a bit position causes the corresponding PBR bit to be compared to the appropriate
PC bit. Setting a PBMR bit causes the corresponding PBR bit to be ignored.
31.4.7
Address Breakpoint Registers (ABLR, ABHR)
The ABLR and ABHR, shown in Figure 31-10 , define regions in the processor’s data address space that
can be used as part of the trigger. These register values are compared with the address for each transfer on
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-15
Debug Module the processor’s high-speed local bus. The trigger definition register (TDR) identifies the trigger as one of three cases:
1. Identical to the value in ABLR
2. Inside the range bound by ABLR and ABHR inclusive
3. Outside that same range
ABHR is accessible in supervisor mode as debug control register 0x0C using the WDEBUG instruction and via the BDM port using the
RDMREG
and
WDMREG
commands. ABLR is accessible in supervisor mode as debug control register 0x0D using the WDEBUG instruction and via the BDM port using the
WDMREG command.
DRc[4:0]: 0x0C (ABHR)
0x0D (ABLR)
Access: Supervisor write-only
BDM write-only
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
R
W Address
Reset – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –
Figure 31-10. Address Breakpoint Registers (ABLR, ABHR)
Table 31-11. ABLR Field Description
Field Description
31–0
Address
Low address. Holds the 32-bit address marking the lower bound of the address breakpoint range. Breakpoints for specific addresses are programmed into ABLR.
Table 31-12. ABHR Field Description
Field Description
31–0
Address
High address. Holds the 32-bit address marking the upper bound of the address breakpoint range.
31.4.8
Data Breakpoint/Mask Registers (DBR, DBMR)
The DBR specifies data patterns used as part of the trigger into debug mode. DBR bits are masked by setting the corresponding DBMR bits, as defined in TDR. DBR is accessible in supervisor mode as debug control register 0x0E, using the WDEBUG instruction, and through the BDM port using the RDMREG and
WDMREG commands. DBMR is accessible in supervisor mode as debug control register 0x0F, using the
WDEBUG instruction and via the BDM port using the WDMREG command.
DRc[4:0]: 0x0E (DBR)
0x0F (DBMR)
Access: Supervisor write-only
BDM write-only
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
R
W Data (DBR); Mask (DBMR)
Reset – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –
Figure 31-11. Data Breakpoint & Mask Registers (DBR & DBMR)
31-16
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
Table 31-13. DBR Field Descriptions
Field
31–0
Data
Description
Data breakpoint value. Contains the value to be compared with the data value from the processor’s local bus as a breakpoint trigger.
Table 31-14. DBMR Field Descriptions
Field
31–0
Mask
Description
Data breakpoint mask. The 32-bit mask for the data breakpoint trigger. Clearing a DBR bit allows the corresponding
DBR bit to be compared to the appropriate bit of the processor’s local data bus. Setting a DBMR bit causes that bit to be ignored.
The DBR supports both aligned and mis-aligned references. Table 31-15
shows relationships between processor address, access size, and location within the 32-bit data bus.
Table 31-15. Access Size and Operand Data Location
A[1:0]
0x
1x xx
00
01
10
11
Access Size
Byte
Byte
Byte
Byte
Word
Word
Longword
Operand Location
D[31:24]
D[23:16]
D[15:8]
D[7:0]
D[31:16]
D[15:0]
D[31:0]
31.5
Background Debug Mode (BDM)
The ColdFire family implements a low-level system debugger in the microprocessor in a dedicated hardware module. Communication with the development system is handled through a dedicated, high-speed serial command interface. Although some BDM operations, such as CPU register accesses, require the CPU to be halted, other BDM commands, such as memory accesses, can be executed while the processor is running.
31.5.1
CPU Halt
Although most BDM operations can occur in parallel with CPU operations, unrestricted BDM operation requires the CPU to be halted. The sources that can cause the CPU to halt are listed below in order of priority:
1. A catastrophic fault-on-fault condition automatically halts the processor.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-17
Debug Module
2. A hardware breakpoint can be configured to generate a pending halt condition similar to the assertion of BKPT. This type of halt is always first made pending in the processor. Next, the processor samples for pending halt and interrupt conditions once per instruction. When a pending
condition is asserted, the processor halts execution at the next sample point. See Section 31.6.1,
3. The execution of a HALT instruction immediately suspends execution. Attempting to execute
HALT in user mode while CSR[UHE] = 0 generates a privilege violation exception. If
CSR[UHE] = 1, HALT can be executed in user mode. After HALT executes, the processor can be restarted by serial shifting a
GO
command into the debug module. Execution continues at the instruction after HALT.
4. The assertion of the BKPT input is treated as a pseudo-interrupt; that is, the halt condition is postponed until the processor core samples for halts/interrupts. The processor samples for these conditions once during the execution of each instruction. If there is a pending halt condition at the sample time, the processor suspends execution and enters the halted state.
The assertion of BKPT should be considered in the following two special cases:
• After the system reset signal is negated, the processor waits for 16 processor clock cycles before beginning reset exception processing. If the BKPT input is asserted within eight cycles after RSTI is negated, the processor enters the halt state, signaling halt status (0xF) on the PST outputs. While the processor is in this state, all resources accessible through the debug module can be referenced.
This is the only chance to force the processor into emulation mode through CSR[EMU].
After system initialization, the processor’s response to the GO command depends on the set of
BDM commands performed while it is halted for a breakpoint. Specifically, if the PC register was loaded, the GO command causes the processor to exit halted state and pass control to the instruction address in the PC, bypassing normal reset exception processing. If the PC was not loaded, the GO command causes the processor to exit halted state and continue reset exception processing.
• The ColdFire architecture also handles a special case of BKPT being asserted while the processor is stopped by execution of the STOP instruction. For this case, the processor exits the stopped mode and enters the halted state, at which point, all BDM commands may be exercised. When restarted, the processor continues by executing the next sequential instruction, that is, the instruction following the STOP opcode.
The CSR[27–24] bits indicate the halt source, showing the highest priority source for multiple halt conditions.
31.5.2
BDM Serial Interface
When the CPU is halted and PST reflects the halt status, the development system can send unrestricted commands to the debug module. The debug module implements a synchronous serial protocol using two inputs (DSCLK and DSI) and one output (DSO), where DSO is specified as a delay relative to the rising edge of the processor clock. See
. The development system serves as the serial communication channel master and must generate DSCLK.
The serial channel operates at a frequency from DC to 1/5 of the PSTCLK frequency. The channel uses full-duplex mode, where data is sent and received simultaneously by both master and slave devices. The
31-18
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module transmission consists of 17-bit packets composed of a status/control bit and a 16-bit data word. As shown
is, DSI is sampled and DSO is driven.
C1 C2 C3 C4
PSTCLK
DSCLK
DSI Current Next
BDM State
Machine
DSO
Current State
Past
Next State
Current
Figure 31-12. BDM Serial Interface Timing
DSCLK and DSI are synchronized inputs. DSCLK acts as a pseudo clock enable and is sampled on the rising edge of the processor clock as well as the DSI. DSO is delayed from the DSCLK-enabled CLK rising edge (registered after a BDM state machine state change). All events in the debug module’s serial state machine are based on the processor clock rising edge. DSCLK must also be sampled low (on a positive edge of CLK) between each bit exchange. The msb is transferred first. Because DSO changes state based on an internally-recognized rising edge of DSCLK, DSO cannot be used to indicate the start of a serial transfer. The development system must count clock cycles in a given transfer. C1–C4 are described as follows:
• C1—First synchronization cycle for DSI (DSCLK is high)
• C2—Second synchronization cycle for DSI (DSCLK is high)
• C3—BDM state machine changes state depending on DSI and whether the entire input data transfer has been transmitted
• C4—DSO changes to next value
NOTE
A not-ready response can be ignored except during a memory-referencing cycle. Otherwise, the debug module can accept a new serial transfer after 32 processor clock periods.
31.5.2.1
Receive Packet Format
The basic receive packet, Figure 31-13 , consists of 16 data bits and 1 status bit.
15
S
15 14 13 12 11 10 9 8 7
Data Field
6
Figure 31-13. Receive BDM Packet
5 4 3 2 1 0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-19
Debug Module
Table 31-16. Receive BDM Packet Field Description
Field
16
S
Description
Status. Indicates the status of CPU-generated messages listed below. The not-ready response can be ignored unless a memory-referencing cycle is in progress. Otherwise, the debug module can accept a new serial transfer after 32 processor clock periods.
S
0
0
1
1
1
Data xxxx
FFFF
0000
0001
FFFF
Message
Valid data transfer
Status OK
Not ready with response; come again
Error–Terminated bus cycle; data invalid
Illegal Command
15–0
Data
Data. Contains the message to be sent from the debug module to the development system. The response message is always a single word, with the data field encoded as shown above.
31.5.2.2
Transmit Packet Format
The basic transmit packet, Figure 31-14 , consists of 16 data bits and 1 reserve bit.
15
—
Field
16
15–0
Data
15 14 13 12 11 10 9 8
Data
7 6
Figure 31-14. Transmit BDM Packet
5 4 3 2
Table 31-17. Transmit BDM Packet Field Description
Description
Reserved, should be cleared.
Data bits 15–0. Contains the data to be sent from the development system to the debug module.
1 0
31.5.3
BDM Command Set
summarizes the BDM command set. Subsequent paragraphs contain detailed descriptions of each command. Issuing a BDM command when the processor is accessing debug module registers using the WDEBUG instruction causes undefined behavior. See
Table 31-20 for register address encodings.
Table 31-18. BDM Command Summary
Command
Read A/D register
Write A/D register
Mnemonic
RAREG
RDREG
WAREG /
WDREG
/
Description
CPU
State
1
Section/
Page
Command
(Hex)
Read the selected address or data register and return the results through the serial interface.
Halted
0x218 {A/D, Reg[2:0]}
Write the data operand to the specified address or data register.
Halted
0x208 {A/D, Reg[2:0]}
31-20
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
Table 31-18. BDM Command Summary (continued)
Command Mnemonic Description
CPU
State
1
Section/
Page
Command
(Hex)
Read memory location
Write memory location
READ
WRITE
Read the data at the memory location specified by the longword address.
Write the operand data to the memory location specified by the longword address.
Steal
0x1900 byte
0x1940 word
0x1980 longword
Steal
0x1800 byte
0x1840 word
0x1880 longword
Steal
0x1D00 byte
0x1D40 word
0x1D80 longword
Dump memory block
Fill memory block
DUMP
FILL
Used with READ to dump large blocks of memory.
An initial READ is executed to set up the starting address of the block and to retrieve the first result. A DUMP command retrieves subsequent operands.
Used with WRITE to fill large blocks of memory.
An initial WRITE is executed to set up the starting address of the block and to supply the first operand. A FILL command writes subsequent operands.
Steal
0x1C00 byte
0x1C40 word
0x1C80 longword
Resume execution
No operation
Synchronize PC to PST/DDATA
Read control register
GO
NOP
RCREG
The pipeline is flushed and refilled before resuming instruction execution at the current PC.
Perform no operation; may be used as a null command.
SYNC _ PC Capture the current PC and display it on the
PST/DDATA outputs
Read the system control register.
Halted
Parallel
Parallel
Halted
0x0C00
0x0000
0x0001
0x2980
Write control register
Read debug module register
WCREG
RDMREG
Write the operand data to the system control register.
Read the debug module register.
Halted
0x2880
Parallel
0x2D {0x4
2
DRc[4:0]}
Write debug module register
WDMREG Write the operand data to the debug module register.
Parallel
1
2
General command effect and/or requirements on CPU operation:
- Halted: The CPU must be halted to perform this command.
- Steal: Command generates bus cycles that can be interleaved with bus accesses.
- Parallel: Command is executed in parallel with CPU activity.
0x4 is a three-bit field.
0x2C {0x4
2
DRc[4:0]}
Unassigned command opcodes are reserved by Freescale. All unused command formats within any revision level perform a
NOP
and return the illegal command response .
31.5.3.1
ColdFire BDM Command Format
All ColdFire family BDM commands include a 16-bit operation word followed by an optional set of one or more extension words, as shown in
.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-21
Debug Module
15 15 14 13
Operation
12 11 10 9
0
8
R/W
7 6
Op Size
Extension Word(s)
Figure 31-15. BDM Command Format
5
0
4
0
3
A/D
2 1
Register
0
Table 31-19. BDM Field Descriptions
Field Description
15–10
Operation
Specifies the command. These values are listed in Table 31-18
.
9
8
R/W
Reserved, should be cleared.
Direction of operand transfer.
0 Data is written to the CPU or to memory from the development system.
1 The transfer is from the CPU to the development system.
7–6
Op Size
Operand data size for sized operations. Addresses are expressed as 32-bit absolute values. Note that a command performing a byte-sized memory read leaves the upper 8 bits of the response data undefined. Referenced data is returned in the lower 8 bits of the response.
00
01
10
11
Operand Size
Byte
Word
Longword
Reserved
Bit Values
8 bits
16 bits
32 bits
—
5–4
3
A/D
Reserved, should be cleared.
Address/data. Determines whether the register field specifies a data or address register.
0 Data register.
1 Address register.
2–0
Register
Contains the register number in commands that operate on processor registers. See
31.5.3.1.1
Extension Words as Required
Some commands require extension words for addresses and/or immediate data. Addresses require two extension words because only absolute long addressing is permitted. Longword accesses are forcibly longword-aligned, and word accesses are forcibly word-aligned. Immediate data can be 1 or 2 words long.
Byte and word data each requires a single extension word, while longword data requires two extension words.
Operands and addresses are transferred most-significant word first. In the following descriptions of the
BDM command set, the optional set of extension words is defined as address, data, or operand data.
31.5.3.2
Command Sequence Diagrams
The command sequence diagram in Figure 31-16 shows serial bus traffic for commands. Each bubble
represents a 17-bit bus transfer. The top half of each bubble indicates the data the development system
31-22
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module sends to the debug module; the bottom half indicates the debug module’s response to the previous development system commands. Command and result transactions overlap to minimize latency.
Commands transmitted to the debug module
Command code transmitted during this cycle
High-order 16 bits of memory address
Low-order 16 bits of memory address
Non-serial-related activity Sequence taken if operation has not completed
READ (LONG)
???
MS ADDR
’NOT READY’
LS ADDR
’NOT READY’
READ
MEMORY
LOCATION
XXX
’NOT READY’
Next
Command
Code
XXX
’ILLEGAL’
NEXT CMD
’NOT READY’
XXX
MS RESULT
NEXT CMD
LS RESULT
Data used from this transfer
XXX
BERR
NEXT CMD
’NOT READY’
Sequence taken if illegal command is received by debug module
Results from previous command
Sequence taken if bus error occurs on memory access
Responses from the debug module
Figure 31-16. Command Sequence Diagram
High- and low-order 16 bits of result
The sequence is as follows:
• In cycle 1, the development system command is issued ( READ in this example). The debug module responds with either the low-order results of the previous command or a command complete status of the previous command, if no results are required.
• In cycle 2, the development system supplies the high-order 16 address bits. The debug module returns a not-ready response unless the received command is decoded as unimplemented, which is indicated by the illegal command encoding. If this occurs, the development system should retransmit the command.
NOTE
A not-ready response can be ignored except during a memory-referencing cycle. Otherwise, the debug module can accept a new serial transfer after 32 processor clock periods.
• In cycle 3, the development system supplies the low-order 16 address bits. The debug module always returns a not-ready response.
• At the completion of cycle 3, the debug module initiates a memory read operation. Any serial transfers that begin during a memory access return a not-ready response.
• Results are returned in the two serial transfer cycles after the memory access completes. For any command performing a byte-sized memory read operation, the upper 8 bits of the response data are undefined and the referenced data is returned in the lower 8 bits. The next command’s opcode is
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-23
Debug Module sent to the debug module during the final transfer. If a memory or register access is terminated with a bus error, the error status (S = 1, DATA = 0x0001) is returned instead of result data.
31.5.3.3
Command Set Descriptions
The following sections describe the commands summarized in
NOTE
The BDM status bit (S) is 0 for normally completed commands; S = 1 for illegal commands, not-ready responses, and transfers with bus-errors.
Section 31.5.2, “BDM Serial Interface
,” describes the receive packet format.
Freescale reserves unassigned command opcodes for future expansion. Unused command formats in any revision level perform a
NOP
and return an illegal command response.
31.5.3.3.1
Read A/D Register ( RAREG / RDREG )
Read the selected address or data register and return the 32-bit result. A bus error response is returned if the CPU core is not halted.
Command/Result Formats:
Command
Result
15
Command Sequence:
14
0x2
13 12 11 10
0x1
9 8 7 6
0x8
5
D[31:16]
D[15:0]
Figure 31-17. RAREG / RDREG Command Format
4
RAREG/RDREG
???
XXX
MS RESULT
NEXT CMD
LS RESULT
3
A/D
2 1
Register
0
Operand Data:
Result Data:
XXX
BERR
NEXT CMD
’NOT READY’
Figure 31-18. RAREG / RDREG Command Sequence
None
The contents of the selected register are returned as a longword value, most-significant word first.
31.5.3.3.2
Write A/D Register ( WAREG / WDREG )
The operand longword data is written to the specified address or data register. A write alters all 32 register bits. A bus error response is returned if the CPU core is not halted.
Command Format:
31-24
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
15 14
0x2
13 12 11 10 9 8 7 6 5 4
0x0 0x8
D[31:16]
D[15:0]
Figure 31-19. WAREG / WDREG Command Format
3
A/D
2 1
Register
0
Command Sequence:
WAREG/WDREG
???
MS DATA
’NOT READY’
LS DATA
’NOT READY’
NEXT CMD
’CMD COMPLETE’
Operand Data:
Result Data:
XXX
BERR
NEXT CMD
’NOT READY’
Figure 31-20. WAREG / WDREG Command Sequence
Longword data is written into the specified address or data register. The data is supplied most-significant word first.
Command complete status is indicated by returning 0xFFFF (with S cleared) when the register write is complete.
31.5.3.3.3
Read Memory Location ( READ )
Read data at the longword address. Address space is defined by BAAR[TT,TM]. Hardware forces low-order address bits to zeros for word and longword accesses to ensure that word addresses are word-aligned and longword addresses are longword-aligned.
Command/Result Formats:
4 3 2
0x0
1 0
Byte
Command
Result
Word Command
X
Result
Longword Command
Result
15 14
0x1
13 12 11 10
0x9
9 8 7 6
0x0
5
A[31:16]
A[15:0]
X X
0x1
X X X X
0x9
X
0x4
A[31:16]
A[15:0]
D[15:0]
0x1 0x9 0x8
A[31:16]
A[15:0]
D[31:16]
D[15:0]
Figure 31-21. READ Command/Result Formats
D[7:0]
0x0
0x0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-25
Debug Module
Command Sequence:
READ (B/W)
???
MS ADDR
’NOT READY’
LS ADDR
’NOT READY’
READ
MEMORY
LOCATION
XXX
’NOT READY’
NEXT CMD
RESULT
XXX
BERR
NEXT CMD
’NOT READY’
READ (LONG)
???
MS ADDR
’NOT READY’
LS ADDR
’NOT READY’
READ
MEMORY
LOCATION
XXX
’NOT READY’
XXX
MS RESULT
NEXT CMD
LS RESULT
Operand Data:
Result Data:
XXX
BERR
NEXT CMD
’NOT READY’
Figure 31-22. READ Command Sequence
The only operand is the longword address of the requested location.
Word results return 16 bits of data; longword results return 32. Bytes are returned in the LSB of a word result; the upper byte is undefined. 0x0001 (S = 1) is returned if a bus error occurs.
31.5.3.3.4
Write Memory Location ( WRITE )
Write data to the memory location specified by the longword address. The address space is defined by
BAAR[TT,TM]. Hardware forces low-order address bits to zeros for word and longword accesses to ensure that word addresses are word-aligned and longword addresses are longword-aligned.
Command Formats:
31-26
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
Byte
Word
Longword
15 14
0x1
13
X X
0x1
X
0x1
12 11 10
0x8
9 8 7 6
0x0
X X X X
A[31:16]
A[15:0]
X
0x8 0x4
A[31:16]
A[15:0]
D[15:0]
0x8 0x8
A[31:16]
A[15:0]
D[31:16]
D[15:0]
Figure 31-23. WRITE Command Format
5 4
D[7:0]
3 2
0x0
1 0
0x0
0x0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-27
Debug Module
Command Sequence:
WRITE (B/W)
???
MS ADDR
’NOT READY’
LS ADDR
’NOT READY’
DATA
’NOT READY’
WRITE
MEMORY
LOCATION
XXX
’NOT READY’
NEXT CMD
’CMD COMPLETE’
XXX
BERR
NEXT CMD
’NOT READY’
WRITE (LONG)
???
MS ADDR
’NOT READY’
LS ADDR
’NOT READY’
MS DATA
’NOT READY’
LS DATA
’NOT READY’
WRITE
MEMORY
LOCATION
XXX
’NOT READY’
NEXT CMD
’CMD COMPLETE’
XXX
BERR
Operand Data:
Result Data:
NEXT CMD
’NOT READY’
Figure 31-24. WRITE Command Sequence
This two-operand instruction requires a longword absolute address that specifies a location to which the data operand is to be written. Byte data is sent as a 16-bit word, justified in the LSB; 16- and 32-bit operands are sent as 16 and 32 bits, respectively.
Command complete status is indicated by returning 0xFFFF (with S cleared) when the register write is complete. A value of 0x0001 (with S set) is returned if a bus error occurs.
31.5.3.3.5
Dump Memory Block ( DUMP )
DUMP is used with the READ command to access large blocks of memory. An initial READ is executed to set up the starting address of the block and to retrieve the first result. If an initial
READ
is not executed before the first DUMP , an illegal command response is returned. The DUMP command retrieves subsequent operands. The initial address is incremented by the operand size (1, 2, or 4) and saved in a temporary register. Subsequent
DUMP
commands use this address, perform the memory read, increment it by the current operand size, and store the updated address in the temporary register.
31-28
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
NOTE
DUMP
does not check for a valid address; it is a valid command only when preceded by
NOP
,
READ
, or another
DUMP
command. Otherwise, an illegal command response is returned.
NOP
can be used for intercommand padding without corrupting the address pointer.
The size field is examined each time a DUMP command is processed, allowing the operand size to be dynamically altered.
Command/Result Formats:
15 12 11 8 7 4 3 0
Byte Command 0x1 0xD 0x0 0x0
X D[7:0]
Word
Result
Command
Result
Longword Command
Result
X X X X X X X
0x1 0xD 0x4
D[15:0]
0x1 0xD 0x8
D[31:16]
D[15:0]
Figure 31-25. DUMP Command/Result Formats
0x0
0x0
Command Sequence:
DUMP (B/W)
???
READ
MEMORY
LOCATION
XXX
’NOT READY’
NEXT CMD
RESULT
XXX
’ILLEGAL’
NEXT CMD
’NOT READY’
XXX
BERR
NEXT CMD
’NOT READY’
DUMP (LONG)
???
Operand Data:
READ
MEMORY
LOCATION
XXX
’NOT READY’
NEXT CMD
MS RESULT
None
XXX
’ILLEGAL’
NEXT CMD
’NOT READY’
Figure 31-26. DUMP Command Sequence
XXX
BERR
NEXT CMD
LS RESULT
NEXT CMD
’NOT READY’
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-29
Debug Module
Result Data:
Byte
Requested data is returned as either a word or longword. Byte data is returned in the least-significant byte of a word result. Word results return 16 bits of significant data; longword results return 32 bits. A value of 0x0001 (with S set) is returned if a bus error occurs.
31.5.3.3.6
Fill Memory Block ( FILL )
A
FILL
command is used with the
WRITE
command to access large blocks of memory. An initial
WRITE
is sets up the starting address of the block and to supply the first operand. The
FILL
command writes subsequent operands. The initial address is incremented by the operand size (1, 2, or 4) and saved in a temporary register after the memory write. Subsequent
FILL
commands use this address, perform the write, increment it by the current operand size, and store the updated address in the temporary register.
If an initial WRITE is not executed preceding the first FILL command, the illegal command response is returned.
NOTE
The FILL command does not check for a valid address— FILL is a valid command only when preceded by another FILL , a NOP , or a WRITE command.
Otherwise, an illegal command response is returned. The NOP command can be used for intercommand padding without corrupting the address pointer.
The size field is examined each time a
FILL
command is processed, allowing the operand size to be altered dynamically.
Command Formats:
4 3 2
0x0
1 0
Word
Longword
15
X
14
0x1
13
X X
0x1
0x1
12
X
11
X
10
0xC
9
X X
0xC
8
X
7 6
0x0
5
0x4
D[15:0]
0xC 0x8
D[31:16]
D[15:0]
Figure 31-27. FILL Command Format
D[7:0]
0x0
0x0
31-30
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
Command Sequence:
FILL (LONG)
???
MS DATA
’NOT READY’
XXX
’ILLEGAL’
LS DATA
’NOT READY’
NEXT CMD
’NOT READY’
WRITE
MEMORY
LOCATION
XXX
’NOT READY’
NEXT CMD
’CMD COMPLETE’
XXX
BERR
NEXT CMD
’NOT READY’
FILL (B/W)
???
DATA
’NOT READY’
WRITE
MEMORY
LOCATION
XXX
’NOT READY’
Operand Data:
Result Data:
XXX
’ILLEGAL’
NEXT CMD
’NOT READY’
NEXT CMD
’CMD COMPLETE’
XXX
BERR
NEXT CMD
’NOT READY’
Figure 31-28. FILL Command Sequence
A single operand is data to be written to the memory location. Byte data is sent as a 16-bit word, justified in the least-significant byte; 16- and 32-bit operands are sent as 16 and 32 bits, respectively.
Command complete status (0xFFFF) is returned when the register write is complete. A value of 0x0001 (with S set) is returned if a bus error occurs.
31.5.3.3.7
Resume Execution ( GO )
The pipeline is flushed and refilled before normal instruction execution resumes. Prefetching begins at the current address in the PC and at the current privilege level. If any register (such as the PC or SR) is altered by a BDM command while the processor is halted, the updated value is used when prefetching resumes.
If a GO command is issued and the CPU is not halted, the command is ignored.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
0x0 0xC 0x0 0x0
Figure 31-29. GO Command Format
Command Sequence:
GO
???
NEXT CMD
’CMD COMPLETE’
Figure 31-30. GO Command Sequence
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-31
Debug Module
Operand Data:
Result Data:
None
The command-complete response (0xFFFF) is returned during the next shift operation.
31.5.3.3.8
No Operation ( NOP )
NOP
performs no operation and may be used as a null command where required.
Command Formats:
15 14
0x0
13 12 11 10 9 8 7 6 5
0x0 0x0
Figure 31-31. NOP Command Format
4 3 2
0x0
1 0
Command Sequence:
NOP
???
NEXT CMD
’CMD COMPLETE’
Operand Data:
Result Data:
Figure 31-32. NOP Command Sequence
None
The command-complete response, 0xFFFF (with S cleared), is returned during the next shift operation.
31.5.3.3.9
Synchronize PC to the PST/DDATA Lines ( SYNC _ PC )
Capture the current PC (program counter) and display it on the PST/DDATA outputs. After the debug module receives the command, it sends a signal to the ColdFire core that the current PC must be displayed.
The core then forces an instruction fetch at the next PC with the address being captured in the DDATA logic under control of the CSR[BTB] bits. The specific sequence of PST and DDATA values is defined below:
1. Debug signals a SYNC _ PC command is pending.
2. CPU completes the current instruction.
3. CPU forces an instruction fetch to the next PC, generates a PST = $5 value indicating a taken branch and signals DDATA. DDATA captures the instruction address corresponding to the PC.
DDATA generates a PST marker ($9–$B) as defined by CSR[BTB] and displays the captured PC address.
This command can be used to dynamically access the PC for performance monitoring. The execution of this command is considerably less obtrusive to the real-time operation of an application than a halt-CPU/read-PC/resume command sequence.
Format:
15 14
0x0
13 12 11 10
0x0
9 8 7 6
0x0
5 4 3 2
0x1
1 0
31-32
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
Command Sequence:
SYNC_PC
???
NEXT CMD
“CMD COMPLETE”
Operand Data:
Result Data:
None
The command complete response, $FFFF (with the status bit cleared), is returned during the next shift operation.
31.5.3.3.10
Read Control Register ( RCREG )
Reads the selected control register and returns the 32-bit result. Accesses to the processor/memory control registers are always 32 bits wide, regardless of register width. The second and third words of the command form a 32-bit address, which the debug module uses to generate a special bus cycle to access the specified control register. The 12-bit Rc field is the same as that used by the MOVEC instruction.
Command/Result Formats:
Command
Result
15 14
0x2
13
0x0
0x0
12 11 10
0x9
9
0x0
8 7 6
0x8
0x0
Rc
5
D[31:16]
D[15:0]
Figure 31-33. RCREG Command/Result Formats
4
Rc encoding:
Rc
0x800
0x801
0x804
0x805
0x806
0x80E
0x80F
0xC04
0xC05
Table 31-20. Control Register Map
Register Definition
Other Stack Pointer (OTHER_A7)
Vector Base Register (VBR)
MAC Status Register (MACSR)
MAC Mask Register (MASK)
MAC Accumulator 0 (ACC0)
Status Register (SR)
Program Register (PC)
Flash Base Address Register (FLASHBAR)
RAM Base Address Register (RAMBAR)
3
Command Sequence:
2
0x0
0x0
1 0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-33
Debug Module
RCREG
???
MS ADDR
’NOT READY’
MS ADDR
’NOT READY’
READ
CONTROL
REGISTER
XXX
’NOT READY’
NEXT CMD
MS RESULT
NEXT CMD
LS RESULT
Operand Data:
Result Data:
XXX
BERR
NEXT CMD
’NOT READY’
Figure 31-34. RCREG Command Sequence
The only operand is the 32-bit Rc control register select field.
Control register contents are returned as a longword, most-significant word first.
The implemented portion of registers smaller than 32 bits is guaranteed correct; other bits are undefined.
BDM Accesses of the Stack Pointer Registers (A7: SSP, USP)
The V2 core supports two unique stack pointer (A7) registers: the supervisor stack pointer (SSP) and the user stack pointer (USP). The hardware implementation of these two programmable-visible, 32-bit registers does not uniquely identify one as the SSP and the other as the USP. Rather, the hardware uses one
32-bit register as the currently-active A7, and the other register is named simply the ‘other_A7.’ Thus, the contents of the two hardware registers is a function of the operating mode of the processor: if SR[S] = 1
then
else
A7 = Supervisor Stack Pointer other_A7 = User Stack Pointer
A7 = User Stack Pointer other_A7 = Supervisor Stack Pointer
The BDM programming model supports reads and writes to the A7 and other_A7 registers directly. It is the responsibility of the external development system to determine the mapping of the two hardware registers (A7, other_A7) to the two program-visible definitions (supervisor and user stack pointers), based on the supervisor bit of the status register.
31.5.3.3.11
Write Control Register ( WCREG )
The operand (longword) data is written to the specified control register. The write alters all 32 register bits.
Command/Result Formats:
31-34
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
Command
Result
15
Command Sequence:
14
0x2
13
0x0
0x0
12 11 10
0x8
9
0x0
8 7 6
0x8
5
0x0
R c
D[31:16]
D[15:0]
Figure 31-35. WCREG Command/Result Formats
4 3 2
0x0
1
0x0
0
WCREG
???
MS ADDR
’NOT READY’
MS ADDR
’NOT READY’
MS DATA
’NOT READY’
LS DATA
’NOT READY’
WRITE
CONTROL
REGISTER
XXX
’NOT READY’
NEXT CMD
’CMD COMPLETE’
XXX
BERR
Operand Data:
Result Data:
NEXT CMD
’NOT READY’
Figure 31-36. WCREG Command Sequence
This instruction requires two longword operands. The first selects the register to which the operand data is to be written; the second contains the data.
Successful write operations return 0xFFFF. Bus errors on the write cycle are indicated by the setting of bit 16 in the status message and by a data pattern of
0x0001.
31.5.3.3.12
Read Debug Module Register ( RDMREG )
Read the selected debug module register and return the 32-bit result. The only valid register selection for the RDMREG command is CSR (DRc = 0x00). Note that this read of the CSR clears CSR[FOF, TRG, HALT,
BKPT]; as well as the trigger status bits (CSR[BSTAT]) if either a level-2 breakpoint has been triggered or a level-1 breakpoint has been triggered and no level-2 breakpoint has been enabled.
Command/Result Formats:
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-35
Debug Module
Command
Result
15 14
0x2
13 12 11 10
0xD
9 8 7 6
100
5 4
D[31:16]
D[15:0]
Figure 31-37. RDMREG Command/Result Formats
3 2
DR c
1
shows the definition of DRc encoding.
Table 31-21. Definition of DRc Encoding—Read
DRc[4:0]
0x00
0x01–0x1F
Debug Register Definition
Configuration/Status
Reserved
Mnemonic
CSR
—
Initial State
0x0
—
Command Sequence:
0
Page
—
RDMREG
???
XXX
MS RESULT
NEXT CMD
LS RESULT
XXX
’ILLEGAL’
NEXT CMD
’NOT READY’
Operand Data:
Result Data:
Figure 31-38. RDMREG Command Sequence
None
The contents of the selected debug register are returned as a longword value. The data is returned most-significant word first.
31.5.3.3.13
Write Debug Module Register ( WDMREG )
The operand (longword) data is written to the specified debug module register. All 32 bits of the register are altered by the write. DSCLK must be inactive while the debug module register writes from the CPU accesses are performed using the WDEBUG instruction.
Command Format:
0 15 14
0x2
13 12 11 10
0xC
9 8 7 6
100
5 4
D[31:16]
D[15:0]
Figure 31-39. WDMREG BDM Command Format
shows the definition of the DRc write encoding.
Command Sequence:
3 2
DRc
1
31-36
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
WDMREG
???
MS DATA
’NOT READY’
LS DATA
’NOT READY’
NEXT CMD
’CMD COMPLETE’
Operand Data:
Result Data:
XXX
’ILLEGAL’
NEXT CMD
’NOT READY’
Figure 31-40. WDMREG Command Sequence
Longword data is written into the specified debug register. The data is supplied most-significant word first.
Command complete status (0xFFFF) is returned when register write is complete.
31.6
Real-Time Debug Support
The ColdFire family provides support debugging real-time applications. For these types of embedded systems, the processor must continue to operate during debug. The foundation of this area of debug support is that while the processor cannot be halted to allow debugging, the system can generally tolerate small intrusions into the real-time operation.
The debug module provides four types of breakpoints—PC with mask, PC without mask, operand address range, and data with mask. These breakpoints can be configured into one- or two-level triggers with the exact trigger response also programmable. The debug module programming model can be written from either the external development system using the debug serial interface or from the processor’s supervisor programming model using the WDEBUG instruction. Only CSR is readable using the external development system.
31.6.1
Theory of Operation
Breakpoint hardware can be configured to respond to triggers in several ways. The desired response is
programmed into TDR. As shown in Table 31-22 , when a breakpoint is triggered, an indication
(CSR[BSTAT]) is provided on the DDATA output port when it is not displaying captured processor status, operands, or branch addresses.
Table 31-22. DDATA[3:0]/CSR[BSTAT] Breakpoint Response
DDATA[3:0]
1
CSR[BSTAT]
1
Breakpoint Status
0000
0010
0100
0000
0001
0010
No breakpoints enabled
Waiting for level-1 breakpoint
Level-1 breakpoint triggered
1010 0101 Waiting for level-2 breakpoint
1100 0110 Level-2 breakpoint triggered
1
Encodings not shown are reserved for future use.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-37
Debug Module
The breakpoint status is also posted in the CSR. Note that CSR[BSTAT] is cleared by a CSR read when either a level-2 breakpoint is triggered, or a level-1 breakpoint is triggered and a level-2 breakpoint is not enabled. Status is also cleared by writing to TDR.
BDM instructions use the appropriate registers to load and configure breakpoints. As the system operates, a breakpoint trigger generates the response defined in TDR.
PC breakpoints are treated in a precise manner—exception recognition and processing are initiated before the excepting instruction is executed. All other breakpoint events are recognized on the processor’s local bus, but are made pending to the processor and sampled like other interrupt conditions. As a result, these interrupts are imprecise.
In systems that tolerate the processor being halted, a BDM-entry can be used. With TDR[TRC] = 01, a breakpoint trigger causes the core to halt (PST = 0xF).
If the processor core cannot be halted, the debug interrupt can be used. With the configuration
TDR[TRC] = 10, the breakpoint trigger becomes a debug interrupt to the processor, which is treated higher than the nonmaskable level-7 interrupt request. As with all interrupts, it is made pending until the processor reaches a sample point, which occurs once per instruction. Again, the hardware forces the PC breakpoint to occur before the targeted instruction executes. This is possible because the PC breakpoint is enabled when interrupt sampling occurs. For address and data breakpoints, reporting is considered imprecise, because several instructions may execute after the triggering address or data is detected.
As soon as the debug interrupt is recognized, the processor aborts execution and initiates exception processing. This event is signaled externally by the assertion of a unique PST value (PST = 0xD) for multiple cycles. The core enters emulator mode when exception processing begins. After the standard
8-byte exception stack is created, the processor fetches a unique exception vector, 12, from the vector table. (Refer to the ColdFire Programmer’s Reference Manual ).
Execution continues at the instruction address in the vector corresponding to the breakpoint triggered. All interrupts are ignored while the processor is in emulator mode. The debug interrupt handler can use supervisor instructions to save the necessary context, such as the state of all program-visible registers into a reserved memory area.
When debug interrupt operations complete, the RTE instruction executes and the processor exits emulator mode. After the debug interrupt handler completes execution, the external development system can use
BDM commands to read the reserved memory locations.
If a hardware breakpoint such as a PC trigger is left unmodified by the debug interrupt service routine, another debug interrupt is generated after the completion of the RTE instruction. Therefore, the debug interrupt service routine should clear the respective TDR setting.
In the Rev. A implementation, if a hardware breakpoint (e.g., a PC trigger) is left unmodified by the debug interrupt service routine, another debug interrupt is generated after the RTE instruction completes execution. In the Rev. B design, the hardware has been modified to inhibit the generation of another debug interrupt during the first instruction after the RTE exits emulator mode. This behavior is consistent with the existing logic involving trace mode, where the execution of the first instruction occurs before another trace exception is generated. This Rev. B enhancement disables all hardware breakpoints until the first instruction after the RTE has completed execution, regardless of the programmed trigger response.
31-38
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
31.6.1.1
Emulator Mode
Emulator mode is used to facilitate non-intrusive emulator functionality. This mode can be entered in three different ways:
• Setting CSR[EMU] forces the processor into emulator mode. EMU is examined only if RSTI is negated and the processor begins reset exception processing. It can be set while the processor is halted before reset exception processing begins. See
• A debug interrupt always puts the processor in emulation mode when debug interrupt exception processing begins.
• Setting CSR[TRC] forces the processor into emulation mode when trace exception processing begins.
While operating in emulation mode, the processor exhibits the following properties:
• All interrupts are ignored, including level-7 interrupts.
• If CSR[MAP] = 1, all caching of memory and the SRAM module are disabled. All memory accesses are forced into a specially mapped address space signaled by TT = 0x2, TM = 0x5 or 0x6.
This includes stack frame writes and the vector fetch for the exception that forced entry into this mode.
The RTE instruction exits emulation mode. The processor status output port provides a unique encoding for emulator mode entry (0xD) and exit (0x7).
31.6.2
Concurrent BDM and Processor Operation
The debug module supports concurrent operation of both the processor and most BDM commands. BDM commands may be executed while the processor is running, except those following operations that access processor/memory registers:
• Read/write address and data registers
• Read/write control registers
For BDM commands that access memory, the debug module requests the processor’s local bus. The processor responds by stalling the instruction fetch pipeline and waiting for current bus activity to complete before freeing the local bus for the debug module to perform its access. After the debug module bus cycle, the processor reclaims the bus.
Breakpoint registers must be carefully configured in a development system if the processor is executing.
The debug module contains no hardware interlocks, so TDR should be disabled while breakpoint registers are loaded, after which TDR can be written to define the exact trigger. This prevents spurious breakpoint triggers.
Because there are no hardware interlocks in the debug unit, no BDM operations are allowed while the CPU is writing the debug’s registers (DSCLK must be inactive).
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-39
Debug Module
31.7
Processor Status, DDATA Definition
This section specifies the ColdFire processor and debug module’s generation of the processor status (PST) and debug data (DDATA) output on an instruction basis. In general, the PST/DDATA output for an instruction is defined as follows:
PST = 0x1, {PST = [0x89B], DDATA= operand} where the {...} definition is optional operand information defined by the setting of the CSR.
The CSR provides capabilities to display operands based on reference type (read, write, or both). A PST value {0x8, 0x9, or 0xB} identifies the size and presence of valid data to follow on the DDATA output {1,
2, or 4 bytes}. Additionally, for certain change-of-flow branch instructions, CSR[BTB] provides the capability to display the target instruction address on the DDATA output {2, 3, or 4 bytes} using a PST value of {0x9, 0xA, or 0xB}.
Instruction asl.l
asr.l
bitrev.l
byterev.l
bcc.{b,w} bchg bchg bclr bclr add.l
add.l
addi.l
addq.l
addx.l
and.l
and.l
andi.l
31.7.1
User Instruction Set
shows the PST/DDATA specification for user-mode instructions. Rn represents any {Dn, An} register. In this definition, the ‘y’ suffix generally denotes the source, and ‘x’ denotes the destination operand. For a given instruction, the optional operand data is displayed only for those effective addresses referencing memory. The ‘DD’ nomenclature refers to the DDATA outputs.
Table 31-23. PST/DDATA Specification for User-Mode Instructions
Operand Syntax
<ea>y,Rx
Dy,<ea>x
#imm,Dx
#imm,<ea>x
Dy,Dx
<ea>y,Dx
Dy,<ea>x
#imm,Dx
{Dy,#imm},Dx
{Dy,#imm},Dx
Dx
Dx
#imm,<ea>x
Dy,<ea>x
#imm,<ea>x
Dy,<ea>x
PST/DDATA
PST = 0x1, {PST = 0xB, DD = source operand}
PST = 0x1, {PST = 0xB, DD = source}, {PST = 0xB, DD = destination}
PST = 0x1
PST = 0x1, {PST = 0xB, DD = source}, {PST = 0xB, DD = destination}
PST = 0x1
PST = 0x1, {PST = 0xB, DD = source operand}
PST = 0x1, {PST = 0xB, DD = source}, {PST = 0xB, DD = destination}
PST = 0x1
PST = 0x1
PST = 0x1
PST = 0x1
PST = 0x1 if taken, then PST = 0x5, else PST = 0x1
PST = 0x1, {PST = 0x8, DD = source}, {PST = 0x8, DD = destination}
PST = 0x1, {PST = 0x8, DD = source}, {PST = 0x8, DD = destination}
PST = 0x1, {PST = 0x8, DD = source}, {PST = 0x8, DD = destination}
PST = 0x1, {PST = 0x8, DD = source}, {PST = 0x8, DD = destination}
31-40
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
lea link.w
lsl.l
lsr.l
move.b
move.l
move.w
move.w
move.w
Instruction clr.w
cmp.l
cmpi.l
divs.l
divs.w
divu.l
divu.w
eor.l
bra.{b,w} bset bset bsr.{b,w} btst btst clr.b
clr.l
eori.l
ext.l
ext.w
extb.l
ff1.l
jmp jsr
Table 31-23. PST/DDATA Specification for User-Mode Instructions (continued)
Operand Syntax PST/DDATA
#imm,<ea>x
Dy,<ea>x
#imm,<ea>x
Dy,<ea>x
<ea>x
<ea>x
<ea>x
<ea>y,Rx
#imm,Dx
<ea>y,Dx
<ea>y,Dx
<ea>y,Dx
<ea>y,Dx
Dy,<ea>x
#imm,Dx
Dx
Dx
Dx
Dx
<ea>x
<ea>x
PST = 0x5
PST = 0x1, {PST = 0x8, DD = source}, {PST = 0x8, DD = destination}
PST = 0x1, {PST = 0x8, DD = source}, {PST = 0x8, DD = destination}
PST = 0x5, {PST = 0xB, DD = destination operand}
PST = 0x1, {PST = 0x8, DD = source operand}
PST = 0x1, {PST = 0x8, DD = source operand}
PST = 0x1, {PST = 0x8, DD = destination operand}
PST = 0x1, {PST = 0xB, DD = destination operand}
PST = 0x1, {PST = 0x9, DD = destination operand}
PST = 0x1, {PST = 0xB, DD = source operand}
PST = 0x1
PST = 0x1, {PST = 0xB, DD = source operand}
PST = 0x1, {PST = 0x9, DD = source operand}
PST = 0x1, {PST = 0xB, DD = source operand}
PST = 0x1, {PST = 0x9, DD = source operand}
PST = 0x1, {PST = 0xB, DD = source}, {PST = 0xB, DD = destination}
PST = 0x1
PST = 0x1
PST = 0x1
PST = 0x1
PST = 0x1
PST = 0x5, {PST = [0x9AB], DD = target address}
1
PST = 0x5, {PST = [0x9AB], DD = target address},
{PST = 0xB , DD = destination operand}
1
<ea>y,Ax
Ay,#imm
{Dy,#imm},Dx
{Dy,#imm},Dx
PST = 0x1
PST = 0x1, {PST = 0xB, DD = destination operand}
PST = 0x1
PST = 0x1
<ea>y,<ea>x
<ea>y,<ea>x
<ea>y,<ea>x
CCR,Dx
PST = 0x1, {PST = 0x8, DD = source}, {PST = 0x8, DD = destination}
PST = 0x1, {PST = 0xB, DD = source}, {PST = 0xB, DD = destination}
PST = 0x1, {PST = 0x9, DD = source}, {PST = 0x9, DD = destination}
PST = 0x1
{Dy,#imm},CCR PST = 0x1
Debug Module
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-41
Debug Module scc sub.l
sub.l
subi.l
subq.l
subx.l
swap trap trapf tst.b
tst.l
tst.w
unlk
Instruction negx.l
nop not.l
or.l
or.l
ori.l
pea pulse movem.l
movem.l
moveq muls.l
muls.w
mulu.l
mulu.w
neg.l
rems.l
remu.l
rts
Table 31-23. PST/DDATA Specification for User-Mode Instructions (continued)
Operand Syntax
#list,<ea>x
<ea>y,#list
PST/DDATA
PST = 0x1, {PST = 0xB, DD = destination},...
2
PST = 0x1, {PST = 0xB, DD = source},...
2
#imm,Dx
<ea>y,Dx
PST = 0x1
PST = 0x1, {PST = 0xB, DD = source operand}
<ea>y,Dx
<ea>y,Dx
<ea>y,Dx
Dx
Dx
PST = 0x1, {PST = 0x9, DD = source operand}
PST = 0x1, {PST = 0xB, DD = source operand}
PST = 0x1, {PST = 0x9, DD = source operand}
PST = 0x1
Dx
<ea>y,Dx
Dy,<ea>x
#imm,Dx
<ea>y
PST = 0x1
PST = 0x1
PST = 0x1
PST = 0x1, {PST = 0xB, DD = source operand}
PST = 0x1, {PST = 0xB, DD = source}, {PST = 0xB, DD = destination}
PST = 0x1
PST = 0x1, {PST = 0xB, DD = destination operand}
PST = 0x4
<ea>y,Dx:Dw PST = 0x1, {PST = 0xB, DD = source operand}
<ea>y,Dx:Dw PST = 0x1, {PST = 0xB, DD = source operand}
PST = 0x1, {PST = 0xB, DD = source operand},
PST = 0x5, {PST = [0x9AB], DD = target address}
Dx
<ea>y,Rx
Dy,<ea>x
#imm,Dx
PST = 0x1
PST = 0x1, {PST = 0xB, DD = source operand}
PST = 0x1, {PST = 0xB, DD = source}, {PST = 0xB, DD = destination}
PST = 0x1
#imm,<ea>x
Dy,Dx
Dx
#imm
<ea>x
<ea>x
<ea>x
Ax
PST = 0x1, {PST = 0xB, DD = source}, {PST = 0xB, DD = destination}
PST = 0x1
PST = 0x1
PST = 0x1
3
PST = 0x1
PST = 0x1, {PST = 0x8, DD = source operand}
PST = 0x1, {PST = 0xB, DD = source operand}
PST = 0x1, {PST = 0x9, DD = source operand}
PST = 0x1, {PST = 0xB, DD = destination operand}
31-42
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Debug Module
Table 31-23. PST/DDATA Specification for User-Mode Instructions (continued)
Instruction Operand Syntax PST/DDATA wddata.b
wddata.l
<ea>y
<ea>y
PST = 0x4, {PST = 0x8, DD = source operand
PST = 0x4, {PST = 0xB, DD = source operand wddata.w
<ea>y PST = 0x4, {PST = 0x9, DD = source operand
1
2
3
For JMP and JSR instructions, the optional target instruction address is displayed only for those effective address fields defining variant addressing modes. This includes the following <ea>x values: (An), (d16,An), (d8,An,Xi),
(d8,PC,Xi).
For Move Multiple instructions (MOVEM), the processor automatically generates line-sized transfers if the operand address reaches a 0-modulo-16 boundary and there are four or more registers to be transferred. For these line-sized transfers, the operand data is never captured nor displayed, regardless of the CSR value.
The automatic line-sized burst transfers are provided to maximize performance during these sequential memory access operations.
During normal exception processing, the PST output is driven to a 0xC indicating the exception processing state. The exception stack write operands, as well as the vector read and target address of the exception handler may also be displayed.
Exception ProcessingPST = 0xC,{PST = 0xB,DD = destination},// stack frame
{PST = 0xB,DD = destination},// stack frame
{PST = 0xB,DD = source},// vector read
PST = 0x5,{PST = [0x9AB],DD = target}// handler PC
The PST/DDATA specification for the reset exception is shown below:
Exception ProcessingPST = 0xC,
The initial references at address 0 and 4 are never captured nor displayed, because these accesses are treated as instruction fetches.
For all types of exception processing, the PST = 0xC value is driven at all times, unless the PST output is needed for one of the optional marker values or for the taken branch indicator (0x5).
shows the PST/DDATA specification for multiply-accumulate instructions.
Table 31-24. PST/DDATA Specification for MAC Instructions
Instruction mac.l
mac.l
mac.w
mac.w
move.l
move.l
move.l
move.l
move.l
move.l
Operand Syntax PST/DDATA
Ry,Rx,Accx PST = 0x1
RyRx,<ea>,Rw,Accx PST = 0x1, {PST = 0xB, DD = source operand}
Ry,Rx,Accx PST = 0x1
Ry,Rx,<ea>,Rw,Accx PST = 0x1, {PST = 0xB, DD = source operand}
<ea>y,Accx PST = 0x1
Accy,Accx PST = 0x1
<ea>y,MACR PST = 0x1
<ea>y,MASK PST = 0x1
<ea>y,Accext01 PST = 0x1
<ea>y,Accext23 PST = 0x1
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-43
Debug Module
Instruction move.l
move.l
move.l
move.l
move.l
move.l
msac.l
msac.l
msac.w
msac.w
Table 31-24. PST/DDATA Specification for MAC Instructions (continued)
Operand Syntax PST/DDATA
Accy,Rx PST = 0x1
MACSR,CCR PST = 0x1
MACSR,Rx PST = 0x1
MASK,Rx PST = 0x1
Accext01,Rx
Accext23,Rx
PST = 0x1
PST = 0x1
Ry,Rx,Accx PST = 0x1
Ry,Rx,<ea>,Rw,Accx PST = 0x1, {PST = 0xB, DD = source operand}
Ry,Rx,Accx PST = 0x1
Ry,Rx,<ea>,Rw,Accx PST = 0x1, {PST = 0xB, DD = source operand}
31.7.2
Supervisor Instruction Set
The supervisor instruction set has complete access to the user mode instructions plus the opcodes shown below. The PST/DDATA specification for these opcodes is shown in
Table 31-25. PST/DDATA Specification for Supervisor-Mode Instructions
Instruction Operand Syntax cpushl halt move.w
move.w
movec rte stldsr.w
stop wdebug
PST/DDATA
SR,Dx PST = 0x1
{Dy,#imm},SR PST = 0x1, {PST = 0x3}
Ry,Rc PST = 0x1
PST = 0x7, {PST = 0xB, DD = source operand}, {PST = 3}, { PST =0xB,
DD =source operand},
PST = 0x5, {[PST = 0x9AB], DD = target address}
#imm
#imm
<ea>y
PST = 0x1
PST = 0x1,
PST = 0xF
PST = 0x1, {PST = 0xA, DD = destination operand, PST = 0x3}
PST = 0x1,
PST = 0xE
PST = 0x1, {PST = 0xB, DD = source, PST = 0xB, DD = source}
The move-to-SR, STLDSR, and RTE instructions include an optional PST = 0x3 value, indicating an entry into user mode. Additionally, if the execution of a RTE instruction returns the processor to emulator mode, a multiple-cycle status of 0xD is signaled.
Similar to the exception processing mode, the stopped state (PST = 0xE) and the halted state (PST = 0xF) display this status throughout the entire time the ColdFire processor is in the given mode.
31.8
Freescale-Recommended BDM Pinout
The ColdFire BDM connector,
, is a 26-pin Berg connector arranged 2 × 13.
31-44
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Developer reserved
1
GND
GND
RSTI
Pad-Voltage
2
GND
PST2
PST0
DDATA2
DDATA0
Freescale reserved
GND
Core-Voltage
10
12
14
16
6
8
2
4
18
20
22
24
26
9
11
13
15
5
7
1
3
17
19
21
23
25
BKPT
DSCLK
Developer reserved
1
DSI
DSO
PST3
PST1
DDATA3
DDATA1
GND
Freescale reserved
PSTCLK
TA
1
Pins reserved for BDM developer use.
2
Supplied by target
Figure 31-41. Recommended BDM Connector
Debug Module
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
31-45
Debug Module
31-46
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Chapter 32
IEEE 1149.1 Test Access Port (JTAG)
32.1
Introduction
The joint test action group (JTAG) is a dedicated user-accessible test logic that complies with the IEEE
1149.1 standard for boundary-scan testability, which helps with system diagnostic and manufacturing testing.
This architecture provides access to all data and chip control pins from the board-edge connector through the standard four-pin test access port (TAP) and the JTAG reset pin, TRST.
32.1.1
Block Diagram
shows the block diagram of the JTAG module.
TDI/DSI
TAP Controller
1-bit Bypass Register
Boundary Scan Register
31
32-bit IDCODE Register
1-bit TEST_CTRL Register
0
0
5-bit TAP Instruction Decoder
4
5-bit TAP Instruction Register
0
1
0
TDO/DSO
JTAG_EN
TCLK
TMS/BKPT
TRST/DSCLK
JTAG Module
Disable DSCLK
Force BKPT = 1
DSI = 0
DSO to Debug Module
Figure 32-1. JTAG Block Diagram
DSI
BKPT
DSCLK
1
0
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
32-1
IEEE 1149.1 Test Access Port (JTAG)
32.1.2
Features
The basic tasks of the JTAG module include the following:
• Performs boundary-scan operations to test circuit board electrical continuity
• Bypasses instruction to reduce the shift register path to a single cell
• Sets chip output pins to safety states while executing the bypass instruction
• Samples the system pins during operation and transparently shift out the result
• Selects between JTAG TAP controller and background debug module (BDM) using a dedicated
JTAG_EN pin
32.1.3
Modes of Operation
The JTAG_EN pin can select between the following modes of operation:
• JTAG mode (JTAG_EN = 1)
•
Background debug mode (BDM)—for more information, refer to Section 31.5, “Background
; (JTAG_EN = 0).
32.2
External Signal Description
The JTAG module has five input and one output external signals, as described in
Table 32-1. Signal Properties
Name
JTAG_EN
TCLK
TMS/BKPT
TDI/DSI
TRST/DSCLK
TDO/DSO
Direction
Input
Input
Input
Input
Input
Output
Function
JTAG/BDM selector input
JTAG Test clock input
JTAG Test mode select / BDM Breakpoint
JTAG Test data input / BDM Development serial input
JTAG Test reset input / BDM Development serial clock
JTAG Test data output / BDM Development serial output
Reset State
—
—
—
—
—
Hi-Z / 0
Pull up
—
Active
Active
Active
Active
—
32.2.1
JTAG Enable (JTAG_EN)
The JTAG_EN pin selects between the debug module and JTAG. If JTAG_EN is low, the debug module is
selected; if it is high, the JTAG is selected. Table 32-2
summarizes the pin function selected depending on
JTAG_EN logic state.
32-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
IEEE 1149.1 Test Access Port (JTAG)
Table 32-2. Pin Function Selected
Module selected
Pin Function
JTAG_EN = 0
BDM
—
BKPT
DSI
DSO
DSCLK
JTAG_EN = 1
JTAG
TCLK
TMS
TDI
TDO
TRST
Pin Name
—
TCLK
BKPT
DSI
DSO
DSCLK
When one module is selected, the inputs into the other module are disabled or forced to a known logic
level, as shown in Table 32-3 , in order to disable the corresponding module.
Table 32-3. Signal State to the Disable Module
Disabling JTAG
Disabling BDM
JTAG_EN = 0
TRST = 0
TMS = 1
—
JTAG_EN = 1
—
Disable DSCLK
DSI = 0
BKPT = 1
NOTE
The JTAG_EN does not support dynamic switching between JTAG and
BDM modes.
32.2.2
Test Clock Input (TCLK)
The TCLK pin is a dedicated JTAG clock input to synchronize the test logic. Pulses on TCLK shift data and instructions into the TDI pin on the rising edge and out of the TDO pin on the falling edge. TCLK is independent of the processor clock. The TCLK pin has an internal pull-up resistor, and holding TCLK high or low for an indefinite period does not cause JTAG test logic to lose state information.
32.2.3
Test Mode Select/Breakpoint (TMS/BKPT)
The TMS pin is the test mode select input that sequences the TAP state machine. TMS is sampled on the rising edge of TCLK. The TMS pin has an internal pull-up resistor.
The BKPT pin is used to request an external breakpoint. Assertion of BKPT puts the processor into a halted state after the current instruction completes.
32.2.4
Test Data Input/Development Serial Input (TDI/DSI)
The TDI pin receives serial test and data, which is sampled on the rising edge of TCLK. Register values are shifted in least significant bit (lsb) first. The TDI pin has an internal pull-up resistor.
The DSI pin provides data input for the debug module serial communication port.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
32-3
IEEE 1149.1 Test Access Port (JTAG)
32.2.5
Test Reset/Development Serial Clock (TRST/DSCLK)
The TRST pin is an active low asynchronous reset input with an internal pull-up resistor that forces the
TAP controller to the test-logic-reset state.
The DSCLK pin clocks the serial communication port to the debug module. Maximum frequency is 1/5 the processor clock speed. At the rising edge of DSCLK, the data input on DSI is sampled and DSO changes state.
32.2.6
Test Data Output/Development Serial Output (TDO/DSO)
The TDO pin is the lsb-first data output. Data is clocked out of TDO on the falling edge of TCLK. TDO is tri-stateable and actively driven in the shift-IR and shift-DR controller states.
The DSO pin provides serial output data in BDM mode.
32.3
Memory Map/Register Definition
The JTAG module registers are not memory mapped and are only accessible through the TDO/DSO pin.
32.3.1
Instruction Shift Register (IR)
The JTAG module uses a -bit shift register with no parity. The IR transfers its value to a parallel hold register and applies an instruction on the falling edge of TCLK when the TAP state machine is in the update-IR state. To load an instruction into the shift portion of the IR, place the serial data on the TDI pin before each rising edge of TCLK. The msb of the IR is the bit closest to the TDI pin, and the lsb is the bit
closest to the TDO pin. See Section 32.4.3, “JTAG Instructions”
for a list of possible instruction codes.
TAP state: Update-IR
R
W
Reset
4
1
3
0
Access: User read/write
2
1
Instruction Code
0
1
0
0 0 0
Figure 32-2. 5-Bit Instruction Register (IR)
0
1
1
32.3.2
IDCODE Register
The IDCODE is a read-only register; its value is chip dependent. For more information, see
Section 32.4.3.1, “IDCODE Instruction
.”
IR[4:0]: 0_0001 (IDCODE) Access: User read-only
R
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
PRN DC PIN JEDEC ID
W
Reset 0 0 0 0 0 1 1 1 0 1 Device Dependent 0 0 0 0 0 0 0 1 1 1 0 1
Figure 32-3. IDCODE Register
32-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
IEEE 1149.1 Test Access Port (JTAG)
Table 32-4. IDCODE Field Descriptions
Field Description
31–28
PRN
27–22
DC
21–12
PIN
Part revision number. Indicate the revision number of the device.
Freescale Design Center number.
Part identification number. Indicate the device number.
11–1
JEDEC
0
ID
Joint Electron Device Engineering Council ID bits. Indicate the reduced JEDEC ID for Freescale (0x0E).
IDCODE register ID. This bit is set to 1 to identify the register as the IDCODE register and not the bypass register according to the IEEE standard 1149.1.
32.3.3
Bypass Register
The bypass register is a single-bit shift register path from TDI to TDO when the BYPASS, CLAMP, or
HIGHZ instructions are selected. After entry into the capture-DR state, the single-bit shift register is set to a logic 0. Therefore, the first bit shifted out after selecting the bypass register is always a logic 0.
32.3.4
TEST_CTRL Register
The TEST_CTRL register is a 1-bit shift register path from TDI to TDO when the
ENABLE_TEST_CTRL instruction is selected. The TEST_CTRL transfers its value to a parallel hold register on the rising edge of TCLK when the TAP state machine is in the update-DR state. The DSE bit selects the drive strength used in JTAG mode.
IR[4:0]: 0_0110 Access: User read-only
R
W
Reset
0
DSE
0
Figure 32-4. 1-Bit TEST_CTRL Register
32.3.5
Boundary Scan Register
The boundary scan register is connected between TDI and TDO when the EXTEST or
SAMPLE/PRELOAD instruction is selected. It captures input pin data, forces fixed values on output pins, and selects a logic value and direction for bidirectional pins, or high impedance for tri-stated pins.
The boundary scan register contains bits for bonded-out and non bonded-out signals, excluding JTAG signals, analog signals, power supplies, compliance enable pins, device configuration pins, and clock signals.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
32-5
IEEE 1149.1 Test Access Port (JTAG)
32.4
Functional Description
32.4.1
JTAG Module
The JTAG module consists of a TAP controller state machine, which is responsible for generating all control signals that execute the JTAG instructions and read/write data registers.
32.4.2
TAP Controller
The TAP controller is a state machine that changes state based on the sequence of logical values on the
TMS signal sampled on the rising edge of the TCLK signal.
Asserting the TRST signal asynchronously resets the TAP controller to the test-logic-reset state. As
shows, holding TMS at logic 1 while clocking TCLK through at least five rising edges also causes the state machine to enter the test-logic-reset state, whatever the initial state.
32-6
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
IEEE 1149.1 Test Access Port (JTAG)
1 TEST-LOGIC-RESET
0
0 RUN-TEST/IDLE
1
SELECT DR-SCAN
1
0
1
CAPTURE-DR
0
SHIFT-DR
1
0
0
1
0
PAUSE-DR
1
EXIT2-DR
1
UPDATE-DR
1 0
0
1
SELECT IR-SCAN
0
1
CAPTURE-IR
0
SHIFT-IR
1
EXIT1-IR
0
0
1
0
0
PAUSE-IR
1
EXIT2-IR
1
UPDATE-IR
1 0
Figure 32-5. TAP Controller State Machine Flow
32.4.3
JTAG Instructions
describes public and private instructions.
Table 32-5. JTAG Instructions
Instruction
IDCODE
SAMPLE/PRELOAD
SAMPLE
IR[4:0] Instruction Summary
00001 Selects IDCODE register for shift
00010 Selects boundary scan register for shifting, sampling, and preloading without disturbing functional operation
00011 Selects boundary scan register for shifting and sampling without disturbing functional operation
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
32-7
IEEE 1149.1 Test Access Port (JTAG)
Table 32-5. JTAG Instructions (continued)
Instruction IR[4:0] Instruction Summary
EXTEST
ENABLE_TEST_CTRL
HIGHZ
CLAMP
00100 Selects boundary scan register while applying preloaded values to output pins and asserting functional reset
00110 Selects TEST_CTRL register
01001 Selects bypass register while tri-stating all output pins and asserting functional reset
01100 Selects bypass while applying fixed values to output pins and asserting functional reset
BYPASS
Reserved
11111 Selects bypass register for data operations all others
1
Decoded to select bypass register
1
Freescale reserves the right to change the decoding of the unused opcodes in the future.
32.4.3.1
IDCODE Instruction
The IDCODE instruction selects the 32-bit IDCODE register for connection as a shift path between the
TDI and TDO pin. This instruction allows interrogation of the MCU to determine its version number and other part identification data. The shift register lsb is forced to logic 1 on the rising edge of TCLK following entry into the capture-DR state. Therefore, the first bit to be shifted out after selecting the
IDCODE register is always a logic 1. The remaining 31 bits are also forced to fixed values on the rising edge of TCLK following entry into the capture-DR state.
IDCODE is the default instruction placed into the instruction register when the TAP resets. Thus, after a
TAP reset, the IDCODE register is selected automatically.
32.4.3.2
SAMPLE/PRELOAD Instruction
The SAMPLE/PRELOAD instruction has two functions:
•
SAMPLE - See Section 32.4.3.3, “SAMPLE Instruction ” for description of this function.
• PRELOAD - initialize the boundary scan register update cells before selecting EXTEST or
CLAMP. This is achieved by ignoring the data shifting out on the TDO pin and shifting in initialization data. The update-DR state and the falling edge of TCLK can then transfer this data to the update cells. The data is applied to the external output pins by the EXTEST or CLAMP instruction.
32.4.3.3
SAMPLE Instruction
The SAMPLE instruction obtains a sample of the system data and control signals present at the MCU input pins and just before the boundary scan cell at the output pins. This sampling occurs on the rising edge of
TCLK in the capture-DR state when the IR contains the 0x2 opcode. The sampled data is accessible by shifting it through the boundary scan register to the TDO output by using the shift-DR state. Both the data capture and the shift operation are transparent to system operation.
32-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
IEEE 1149.1 Test Access Port (JTAG)
NOTE
External synchronization is required to achieve meaningful results because there is no internal synchronization between TCLK and the system clock.
32.4.3.4
EXTEST Instruction
The external test (EXTEST) instruction selects the boundary scan register. It forces all output pins and bidirectional pins configured as outputs to the values preloaded with the SAMPLE/PRELOAD instruction and held in the boundary scan update registers. EXTEST can also configure the direction of bidirectional pins and establish high-impedance states on some pins. EXTEST asserts internal reset for the MCU system logic to force a predictable internal state while performing external boundary scan operations.
32.4.3.5
ENABLE_TEST_CTRL Instruction
The ENABLE_TEST_CTRL instruction selects a 1-bit shift register (TEST_CTRL) for connection as a shift path between the TDI and TDO pin. When the user transitions the TAP controller to the UPDATE_DR state, the register transfers its value to a parallel hold register.
32.4.3.6
HIGHZ Instruction
The HIGHZ instruction eliminates the need to backdrive the output pins during circuit-board testing.
HIGHZ turns off all output drivers, including the 2-state drivers, and selects the bypass register. HIGHZ also asserts internal reset for the MCU system logic to force a predictable internal state.
32.4.3.7
CLAMP Instruction
The CLAMP instruction selects the 1-bit bypass register and asserts internal reset while simultaneously forcing all output pins and bidirectional pins configured as outputs to the fixed values that are preloaded and held in the boundary scan update register. CLAMP enhances test efficiency by reducing the overall shift path to a single bit (the bypass register) while conducting an EXTEST type of instruction through the boundary scan register.
32.4.3.8
BYPASS Instruction
The BYPASS instruction selects the bypass register, creating a single-bit shift register path from the TDI pin to the TDO pin. BYPASS enhances test efficiency by reducing the overall shift path when a device other than the ColdFire processor is the device under test on a board design with multiple chips on the overall boundary scan chain. The shift register lsb is forced to logic 0 on the rising edge of TCLK after entry into the capture-DR state. Therefore, the first bit shifted out after selecting the bypass register is always logic 0. This differentiates parts that support an IDCODE register from parts that support only the bypass register.
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
32-9
IEEE 1149.1 Test Access Port (JTAG)
32.5
Initialization/Application Information
32.5.1
Restrictions
The test logic is a static logic design, and TCLK can be stopped in either a high or low state without loss of data. However, the system clock is not synchronized to TCLK internally. Any mixed operation using both the test logic and the system functional logic requires external synchronization.
Using the EXTEST instruction requires a circuit-board test environment that avoids device-destructive configurations in which MCU output drivers are enabled into actively driven networks.
Low-power stop mode considerations:
• The TAP controller must be in the test-logic-reset state to either enter or remain in the low-power stop mode. Leaving the test-logic-reset state negates the ability to achieve low-power, but does not otherwise affect device functionality.
• The TCLK input is not blocked in low-power stop mode. To consume minimal power, the TCLK input should be externally connected to EV
DD
.
• The TMS, TDI, and TRST pins include on-chip pull-up resistors. For minimal power consumption in low-power stop mode, these three pins should be either connected to EV
DD
or left unconnected.
32.5.2
Nonscan Chain Operation
Keeping the TAP controller in the test-logic-reset state ensures that the scan chain test logic is transparent to the system logic. It is recommended that TMS, TDI, TCLK, and TRST be pulled up. TRST could be connected to ground. However, since there is a pull-up on TRST, some amount of current results. The internal power-on reset input initializes the TAP controller to the test-logic-reset state on power-up without asserting TRST.
32-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Appendix A
Register Memory Map Quick Reference
summarizes the address, name, and byte assignment for registers within the MCF52235 CPU
space. Table A-2 lists an overview of the memory map for the on-chip modules, and
memory map including all of the registers for on-chip modules.
Table A-1. CPU Space Register Memory Map
Address
CPU @ 0x002
CPU @ 0x004
CPU @ 0x005
CPU @ 0x800
CPU @ 0x801
CPU @ 0x804
CPU @ 0x805
CPU @ 0x806
CPU @ 0x80E
CPU @ 0x80F
CPU @ 0xC04
CPU @ 0xC05
Name
Cache Control Register
Access Control Register 0
Access Control Register 1
Other Stack Pointer
Vector Base Register
MAC Status Register
MAC Mask Register
MAC Accumulator 0
Status Register
Program Counter
Flash Base Address Register
RAM Base Address Register
Mnemonic
CACR
ACR0
ACR1
OTHER_A7
VBR
MACSR
MASK
ACC0
SR
PC
FLASHBAR
RAMBAR
Size
32
8
16
16
32
32
32
32
16
32
32
32
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
A-1
Register Memory Map Quick Reference
Table A-2. Module Memory Map Overview
Address Module
0x0000_0000
IPSBAR + 0x00_0000
IPSBAR + 0x00_0040
IPSBAR + 0x00_0080
IPSBAR + 0x00_0100
IPSBAR + 0x00_0110
IPSBAR + 0x00_0120
IPSBAR + 0x00_0130
IPSBAR + 0x00_0140
IPSBAR + 0x00_0200
IPSBAR + 0x00_0240
IPSBAR + 0x00_0280
IPSBAR + 0x00_02C0
IPSBAR + 0x00_0300
IPSBAR + 0x00_0340
On-chip Flash/RAM Array
System Control Module
Reserved
Reserved
DMA (Channel 0)
DMA (Channel 1)
DMA (Channel 2)
DMA (Channel 3)
Reserved
UART0
UART1
UART2
Reserved
I
2
C
QSPI
IPSBAR + 0x00_0380
IPSBAR + 0x00_03C0
IPSBAR + 0x00_0400
IPSBAR + 0x00_0440
IPSBAR + 0x00_0480
IPSBAR + 0x00_04C0
IPSBAR + 0x00_0500
IPSBAR + 0x00_0C00
Reserved
RTC
DMA Timer 0
DMA Timer 1
DMA Timer 2
DMA Timer 3
Reserved
Interrupt Controller 0
IPSBAR + 0x00_0D00
IPSBAR + 0x00_0E00
IPSBAR + 0x00_0F00
IPSBAR + 0x00_1000
Interrupt Controller 1
Reserved
Global Interrupt Acknowledge Cycles
FEC Registers and MIB RAM
IPSBAR + 0x00_1400
IPSBAR + 0x00_1800
FEC FIFO Memory
Reserved
IPSBAR + 0x10_0000 Ports
IPSBAR + 0x11_0000 Reset Controller, Chip Configuration, and Power Management
IPSBAR + 0x12_0000
IPSBAR + 0x13_0000
IPSBAR + 0x14_0000
IPSBAR + 0x15_0000
Clock Module
Edge Port 0
Edge Port 1
Programmable Interval Timer 0
Size
64 bytes
64 bytes
64 bytes
64 bytes
64 bytes
64 bytes
1792
256 bytes
256 bytes
256
256 bytes
1K
1K
1M - 6K
64K
64K
64K
64K
64K
64K
1G
64 bytes
64 bytes
128 bytes
16 bytes
16 bytes
16 bytes
16 bytes
196 bytes
64 bytes
64 bytes
64 bytes
64 bytes
64 bytes
64 bytes
A-2
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Address
IPSBAR + 0x000
IPSBAR + 0x008
IPSBAR + 0x00C
IPSBAR + 0x010
IPSBAR + 0x011
IPSBAR + 0x012
IPSBAR + 0x013
IPSBAR + 0x014
IPSBAR + 0x01C
IPSBAR + 0x020
IPSBAR + 0x024
IPSBAR + 0x025
IPSBAR + 0x026
IPSBAR + 0x027
IPSBAR + 0x028
IPSBAR + 0x8000_0000 Reserved
Register Memory Map Quick Reference
Table A-2. Module Memory Map Overview (continued)
Address
IPSBAR + 0x16_0000
IPSBAR + 0x17_0000
IPSBAR + 0x18_0000
IPSBAR + 0x19_0000
IPSBAR + 0x1A_0000
IPSBAR + 0x1B_0000
IPSBAR + 0x1C_0000
IPSBAR + 0x1D_0000
IPSBAR + 0x1E_0000
IPSBAR + 0x1F_0000
IPSBAR + 0x20_0000
IPSBAR + 0x400_0000
IPSBAR + 0x0408_0000
Module
Programmable Interval Timer 1
Reserved
Reserved
ADC
General Purpose Timer A
PWM
FlexCAN
CFM (Flash) Control Registers
Ethernet Physical Transceiver
Random Number Generator H/W Accelerator
Reserved
CFM (Flash) Memory for IPS Reads and Writes
Reserved
Size
64K
64K
64K
62M
256K
1G - 64M -
256K
2G
64K
64K
64K
64K
64K
64K
64K
Table A-3. Register Memory Map
Name
SCM Registers
Internal Peripheral System Base Address Register
Peripheral Power Management Register - High
Peripheral Power Management Register - Low
Core Reset Status Register
Core Watchdog Control Register
Low-Power Interrupt Control Register
Core Watchdog Service Register
DMA Request Control Register
Default Bus Master Park Register
Master Privilege Register
Peripheral Access Control Register 0
Peripheral Access Control Register 1
Peripheral Access Control Register 2
Peripheral Access Control Register 3
Peripheral Access Control Register 4
Mnemonic
IPSBAR
PPMRH
PPMRL
CRSR
CWCR
LPICR
CWSR
DMAREQC
MPARK
MPR
PACR0
PACR1
PACR2
PACR3
PACR4
Size
8
8
8
8
8
32
8
8
8
8
32
32
32
32
8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
A-3
Register Memory Map Quick Reference
Address
IPSBAR + 0x02A
IPSBAR + 0x02B
IPSBAR + 0x02C
IPSBAR + 0x02E
IPSBAR + 0x030
IPSBAR + 0x031
IPSBAR + 0x100
IPSBAR + 0x104
IPSBAR + 0x108
IPSBAR + 0x10C
IPSBAR + 0x120
IPSBAR + 0x124
IPSBAR + 0x128
IPSBAR + 0x12C
IPSBAR + 0x130
IPSBAR + 0x134
IPSBAR + 0x138
IPSBAR + 0x13C
IPSBAR + 0x140
IPSBAR + 0x144
IPSBAR + 0x148
IPSBAR + 0x14C
IPSBAR + 0x200
IPSBAR + 0x204
IPSBAR + 0x208
IPSBAR + 0x20C
Table A-3. Register Memory Map (continued)
Name
Peripheral Access Control Register 5
Peripheral Access Control Register 6
Peripheral Access Control Register 7
Peripheral Access Control Register 8
Grouped Peripheral Access Control Register 0
Grouped Peripheral Access Control Register 1
DMA Registers
Source Address Register 0
Destination Address Register 0
Byte Count Register 0
DMA Status Register 0
Source Address Register 1
Destination Address Register 1
Byte Count Register 1
DMA Status Register 1
Source Address Register 2
Destination Address Register 2
Byte Count Register 2
DMA Status Register 2
Source Address Register 3
Destination Address Register 3
Byte Count Register 3
DMA Status Register 3
UART Registers
UART Mode Register 0
1
Mnemonic
PACR5
PACR6
PACR7
PACR8
GPACR0
GPACR1
SAR2
DAR2
BCR2
DSR2
SAR3
DAR3
BCR3
DSR3
SAR0
DAR0
BCR0
DSR0
SAR1
DAR1
BCR1
DSR1
UMR10,
UMR20
USR0
UCSR0
(Read) UART Status Register 0
(Write) UART Clock Select Register 0
1
(Read) Reserved
(Write) UART Command Register 0
(Read) UART Receive Buffer 0
(Write) UART Transmit Buffer 0
UCR0
URB0
UTB0
Size
8
8
8
8
8
8
32
32
32
8
32
32
32
8
32
32
32
8
32
32
32
8
8
8
8
8
8
8
8
A-4
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Address
IPSBAR + 0x210
IPSBAR + 0x214
IPSBAR + 0x218
IPSBAR + 0x21C
IPSBAR + 0x234
IPSBAR + 0x238
IPSBAR + 0x23C
IPSBAR + 0x240
IPSBAR + 0x244
IPSBAR + 0x248
IPSBAR + 0x24C
IPSBAR + 0x250
IPSBAR + 0x254
IPSBAR + 0x258
IPSBAR + 0x25C
IPSBAR + 0x274
Register Memory Map Quick Reference
Table A-3. Register Memory Map (continued)
Name
(Read) UART Input Port Change Register 0
(Write) UART Auxiliary Control Register 0
1
(Read) UART Interrupt Status Register 0
(Write) UART Interrupt Mask Register 0
(Read) Reserved
UART Baud Rate Generator Register 10
(Read) Reserved
UART Baud Rate Generator Register 20
(Read) UART Input Port Register 0
(Write) Reserved
(Read) Reserved
(Write) UART Output Port Bit Set Command Register 0
(Read) Reserved
(Write) UART Output Port Bit Reset Command Register 0
UART Mode Registers 1
1
Mnemonic
UIPCR0
UACR0
UISR0
UIMR0
UBG10
UBG20
UIP0
UOP10
UIP00
UMR11,
UMR21
USR1
UCSR1
(Read) UART Status Register 1
(Write) UART Clock Select Register 1
1
(Read) Reserved
(Write) UART Command Register 1
(UART/Read) UART Receive Buffer 1
(UART/Write) UART Transmit Buffer 1
(Read) UART Input Port Change Register 1
(Write) UART Auxiliary Control Register 1
1
(Read) UART Interrupt Status Register 1
(Write) UART Interrupt Mask Register 1
(Read) Reserved
UART Baud Rate Generator Register 11
(Read) Reserved
UART Baud Rate Generator Register 21
(Read) UART Input Port Register 1
(Write) Reserved
UCR1
URB1
UTB1
UIPCR1
UACR1
UISR1
UIMR1
UBG11
UBG21
UIP1 8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
Size
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
A-5
Register Memory Map Quick Reference
Address
IPSBAR + 0x278
IPSBAR + 0x27C
IPSBAR + 0x280
IPSBAR + 0x284
IPSBAR + 0x288
IPSBAR + 0x28C
IPSBAR + 0x290
IPSBAR + 0x294
IPSBAR + 0x298
IPSBAR + 0x29C
IPSBAR + 0x2B4
IPSBAR + 0x2B8
IPSBAR + 0x2BC
IPSBAR + 0x300
IPSBAR + 0x304
IPSBAR + 0x308
IPSBAR + 0x30C
IPSBAR + 0x310
Table A-3. Register Memory Map (continued)
Name
(Read) Reserved
(Write) UART Output Port Bit Set Command Register 1
(Read) Reserved
(Write) UART Output Port Bit Reset Command Register 1
UART Mode Register 2
1
Mnemonic
UOP11
UIP01
UMR12,
UMR22
USR2
UCSR2
(Read) UART Status Register 2
(Write) UART Clock Select Register 2
1
(Read) Reserved
(Write) UART Command Register 2
(Read) UART Receive Buffer 2
(Write) UART Transmit Buffer 2
(Read) UART Input Port Change Register 2
(Write) UART Auxiliary Control Register 2
1
(Read) UART Interrupt Status Register 2
(Write) UART Interrupt Mask Register 2
(Read) Reserved
UART Baud Rate Generator Register 12
(Read) Reserved
UART Baud Rate Generator Register 22
(Read) UART Input Port Register 2
(Write) Reserved
(Read) Reserved
(Write) UART Output Port Bit Set Command Register 2
(Read) Reserved
(Write) UART Output Port Bit Reset Command Register 2
I
2
C Registers
I
2
C Address Register
I
2
C Frequency Divider Register
I
2
C Control Register
I
2
C Status Register
I
2
C Data I/O Register
QSPI Registers
UCR2
URB2
UTB2
UIPCR2
UACR2
UISR2
UIMR2
UBG12
UBG22
UIP2
UOP12
UIP02
I2ADR
I2FDR
I2CR
I2SR
I2DR
A-6
8
8
8
8
8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Size
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
Address
IPSBAR + 0x340
IPSBAR + 0x344
IPSBAR + 0x348
IPSBAR + 0x34C
IPSBAR + 0x350
IPSBAR + 0x354
IPSBAR + 0x3C0
IPSBAR + 0x3C4
IPSBAR + 0x3C8
IPSBAR + 0x3CC
IPSBAR + 0x3D0
IPSBAR + 0x3D4
IPSBAR + 0x3D8
IPSBAR + 0x3DC
IPSBAR + 0x3E0
IPSBAR + 0x3E4
IPSBAR + 0x400
IPSBAR + 0x402
IPSBAR + 0x403
IPSBAR + 0x404
IPSBAR + 0x408
IPSBAR + 0x40C
IPSBAR + 0x440
IPSBAR + 0x442
IPSBAR + 0x443
IPSBAR + 0x444
IPSBAR + 0x448
IPSBAR + 0x44C
IPSBAR + 0x480
IPSBAR + 0x482
Table A-3. Register Memory Map (continued)
Name
QSPI Mode Register
QSPI Delay Register
QSPI Wrap Register
QSPI Interrupt Register
QSPI Address Register
QSPI Data Register
Real Time Clock Registers
RTC Hours and Minutes Register
RTC Seconds Counter Register
RTC Hours and Minutes Alarm Register
RTC Seconds Alarm Register
RTC Control Register
RTC Interrupt Status Register
RTC Interrupt Enable Register
RTC Stopwatch Minutes Register
RTC Days Counter Register
RTC Days Alarm Register
DMA Timer Registers
DMA Timer Mode Register 0
DMA Timer Extended Mode Register 0
DMA Timer Event Register 0
DMA Timer Reference Register 0
DMA Timer Capture Register 0
DMA Timer Counter Register 0
DMA Timer Mode Register 1
DMA Timer Extended Mode Register 1
DMA Timer Event Register 1
DMA Timer Reference Register 1
DMA Timer Capture Register 1
DMA Timer Counter Register 1
DMA Timer Mode Register 2
DMA Timer Extended Mode Register 2
Register Memory Map Quick Reference
Mnemonic
QMR
QDLYR
QWR
QIR
QAR
QDR
DTMR0
DTXMR0
DTER0
DTRR0
DTCR0
DTCN0
DTMR1
DTXMR1
DTER1
DTRR1
DTCR1
DTCN1
DTMR2
DTXMR2
HOURMIN
SECONDS
ALARM_HM
ALARM_SEC
RTCCTL
RTCISR
RTCIENR
STPWCH
DAYS
ALARM_DAY
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
8
32
32
32
16
8
32
32
16
8
16
8
8
32
Size
16
16
16
16
16
16
32
32
32
32
32
32
32
32
32
32
A-7
Address
IPSBAR + 0x483
IPSBAR + 0x484
IPSBAR + 0x488
IPSBAR + 0x48C
IPSBAR + 0x4C0
IPSBAR + 0x4C2
IPSBAR + 0x4C3
IPSBAR + 0x4C4
IPSBAR + 0x4C8
IPSBAR + 0x4CC
IPSBAR + 0xC00
IPSBAR + 0xC04
IPSBAR + 0xC08
IPSBAR + 0xC0C
IPSBAR + 0xC10
IPSBAR + 0xC14
IPSBAR + 0xC18
IPSBAR + 0XC19
IPSBAR + 0xC41
IPSBAR + 0xC42
IPSBAR + 0xC43
IPSBAR + 0xC44
IPSBAR + 0xC45
IPSBAR + 0xC46
IPSBAR + 0xC47
IPSBAR + 0xC48
IPSBAR + 0xC49
IPSBAR + 0xC4A
IPSBAR + 0xC4B
IPSBAR + 0xC4C
IPSBAR + 0xC4D
Register Memory Map Quick Reference
Table A-3. Register Memory Map (continued)
Name
DMA Timer Event Register 2
DMA Timer Reference Register 2
DMA Timer Capture Register 2
DMA Timer Counter Register 2
DMA Timer Mode Register 3
DMA Timer Extended Mode Register 3
DMA Timer Event Register 3
DMA Timer Reference Register 3
DMA Timer Capture Register 3
DMA Timer Counter Register 3
Interrupt Controller 0
Interrupt Pending Register High 0
Interrupt Pending Register Low 0
Interrupt Mask Register High 0
Interrupt Mask Register Low 0
Interrupt Force Register High 0
Interrupt Force Register Low 0
Interrupt Request Level Register 0
Interrupt Acknowledge Level and Priority Register 0
Interrupt Control Register 0-01
Interrupt Control Register 0-02
Interrupt Control Register 0-03
Interrupt Control Register 0-04
Interrupt Control Register 0-05
Interrupt Control Register 0-06
Interrupt Control Register 0-07
Interrupt Control Register 0-08
Interrupt Control Register 0-09
Interrupt Control Register 0-10
Interrupt Control Register 0-11
Interrupt Control Register 0-12
Interrupt Control Register 0-13
ICR001
ICR002
ICR003
ICR004
ICR005
ICR006
ICR007
ICR008
IPRH0
IPRL0
IMRH0
IMRL0
INTFRCH0
INTFRCL0
IRLR0
IACKLPR0
ICR009
ICR010
ICR011
ICR012
ICR013
Mnemonic
DTER2
DTRR2
DTCR2
DTCN2
DTMR3
DTXMR3
DTER3
DTRR3
DTCR3
DTCN3
Size
16
8
8
32
8
32
32
32
32
32
8
8
8
8
8
8
8
8
8
8
8
8
8
32
32
8
8
32
32
32
32
A-8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Table A-3. Register Memory Map (continued)
Name
Interrupt Control Register 0-14
Interrupt Control Register 0-15
Interrupt Control Register 0-17
Interrupt Control Register 0-18
Interrupt Control Register 0-19
Interrupt Control Register 0-20
Interrupt Control Register 0-21
Interrupt Control Register 0-22
Interrupt Control Register 0-23
Interrupt Control Register 0-24
Interrupt Control Register 0-25
Interrupt Control Register 0-26
Interrupt Control Register 0-27
Interrupt Control Register 0-28
Interrupt Control Register 0-29
Interrupt Control Register 0-30
Interrupt Control Register 0-31
Interrupt Control Register 0-32
Interrupt Control Register 0-33
Interrupt Control Register 0-34
Interrupt Control Register 0-35
Interrupt Control Register 0-36
Interrupt Control Register 0-37
Interrupt Control Register 0-38
Interrupt Control Register 0-39
Interrupt Control Register 0-40
Interrupt Control Register 0-41
Interrupt Control Register 0-42
Interrupt Control Register 0-43
Interrupt Control Register 0-44
Interrupt Control Register 0-45
Interrupt Control Register 0-46
Interrupt Control Register 0-47
Register Memory Map Quick Reference
Address
IPSBAR + 0xC5F
IPSBAR + 0xC60
IPSBAR + 0xC61
ISPBAR + 0xC62
IPSBAR + 0xC63
IPSBAR + 0xC64
IPSBAR + 0xC65
IPSBAR + 0xC66
IPSBAR + 0xC67
IPSBAR + 0xC68
IPSBAR + 0xC69
IPSBAR + 0xC6A
IPSBAR + 0xC6B
IPSBAR + 0xC6C
IPSBAR + 0xC6D
IPSBAR + 0xC6E
IPSBAR + 0xC6F
IPSBAR + 0xC4E
IPSBAR + 0xC4F
IPSBAR + 0xC51
IPSBAR + 0xC52
IPSBAR +0xC53
IPSBAR + 0xC54
IPSBAR + 0xC55
IPSBAR + 0xC56
IPSBAR + 0xC57
IPSBAR + 0xC58
IPSBAR + 0xC59
IPSBAR + 0xC5A
IPSBAR + 0xC5B
IPSBAR + 0xC5C
IPSBAR + 0xC5D
IPSBAR + 0xC5E
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Size
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
Mnemonic
ICR039
ICR040
ICR041
ICR042
ICR043
ICR044
ICR045
ICR046
ICR047
ICR031
ICR032
ICR033
ICR034
ICR035
ICR036
ICR037
ICR038
ICR023
ICR024
ICR025
ICR026
ICR027
ICR028
ICR029
ICR030
ICR014
ICR015
ICR017
ICR018
ICR019
ICR020
ICR021
ICR022
A-9
Address
IPSBAR + 0xC70
IPSBAR + 0xC71
IPSBAR + 0xC72
IPSBAR + 0xC73
IPSBAR + 0xC74
IPSBAR + 0xC75
IPSBAR + 0xC76
IPSBAR + 0xC77
IPSBAR + 0xC78
IPSBAR + 0xC79
IPSBAR + 0xC7A
IPSBAR + 0xC7B
IPSBAR + 0xC7C
IPSBAR + 0xC7D
IPSBAR + 0xC7E
IPSBAR + 0xCE0
IPSBAR + 0xCE4
IPSBAR + 0xCE8
IPSBAR + 0xCEC
IPSBAR + 0xCF0
IPSBAR + 0xCF4
IPSBAR + 0xCF8
IPSBAR + 0xCFC
Register Memory Map Quick Reference
IPSBAR + 0xD00
IPSBAR + 0xD04
IPSBAR + 0xD08
IPSBAR + 0xD0C
IPSBAR + 0xD10
IPSBAR + 0xD14
IPSBAR + 0xD18
IPSBAR + 0xD19
Table A-3. Register Memory Map (continued)
Name
Interrupt Control Register 0-48
Interrupt Control Register 0-49
Interrupt Control Register 0-50
Interrupt Control Register 0-51
Interrupt Control Register 0-52
Interrupt Control Register 0-53
Interrupt Control Register 0-54
Interrupt Control Register 0-55
Interrupt Control Register 0-56
Interrupt Control Register 0-57
Interrupt Control Register 0-58
Interrupt Control Register 0-59
Interrupt Control Register 0-60
Interrupt Control Register 0-61
Interrupt Control Register 0-62
Software Interrupt Acknowledge Register 0
Level 1 Interrupt Acknowledge Register 0
Level 2 Interrupt Acknowledge Register 0
Level 3 Interrupt Acknowledge Register 0
Level 4 Interrupt Acknowledge Register 0
Level 5 Interrupt Acknowledge Register 0
Level 6 Interrupt Acknowledge Register 0
Level 7 Interrupt Acknowledge Register 0
Interrupt Controller 1
Interrupt Pending Register High 11
Interrupt Pending Register Low 11
Interrupt Mask Register High 11
Interrupt Mask Register Low 1
Interrupt Force Register High 1
Interrupt Force Register Low 1
Interrupt Request Level Register 1
Interrupt Acknowledge Level and Priority Register 1
Mnemonic
ICR056
ICR057
ICR058
ICR059
ICR060
ICR061
ICR062
SWACKR0
ICR048
ICR049
ICR050
ICR051
ICR052
ICR053
ICR054
ICR055
L1IACKR0
L2IACKR0
L3IACKR0
L4IACKR0
L5IACKR0
L6IACKR0
L7IACKR0
IPRH1
IPRL1
IMRH1
IMRL1
INTFRCH1
INTFRCL1
IRLR1
IACKLPR1
Size
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
32
32
8
8
32
32
32
32
A-10
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Table A-3. Register Memory Map (continued)
Name
Interrupt Control Register 1-01
Interrupt Control Register 1-02
Interrupt Control Register 1-03
Interrupt Control Register 1-04
Interrupt Control Register 1-05
Interrupt Control Register 1-06
Interrupt Control Register 1-07
Interrupt Control Register 1-08
Interrupt Control Register 1-09
Interrupt Control Register 1-10
Interrupt Control Register 1-11
Interrupt Control Register 1-12
Interrupt Control Register 1-13
Interrupt Control Register 1-14
Interrupt Control Register 1-15
Interrupt Control Register 1-17
Interrupt Control Register 1-18
Interrupt Control Register 1-19
Interrupt Control Register 1-20
Interrupt Control Register 1-21
Interrupt Control Register 1-22
Interrupt Control Register 1-23
Interrupt Control Register 1-24
Interrupt Control Register 1-25
Interrupt Control Register 1-26
Interrupt Control Register 1-27
Interrupt Control Register 1-28
Interrupt Control Register 1-29
Interrupt Control Register 1-30
Interrupt Control Register 1-31
Interrupt Control Register 1-32
Interrupt Control Register 1-33
Interrupt Control Register 1-34
Register Memory Map Quick Reference
Address
IPSBAR + 0xD52
IPSBAR +0xC53
IPSBAR + 0xD54
IPSBAR + 0xD55
IPSBAR + 0xD56
IPSBAR + 0xD57
IPSBAR + 0xD58
IPSBAR + 0xD59
IPSBAR + 0xD5A
IPSBAR + 0xD5B
IPSBAR + 0xD5C
IPSBAR + 0xD5D
IPSBAR + 0xD5E
IPSBAR + 0xD5F
IPSBAR + 0xD61
IPSBAR + 0xD61
ISPBAR + 0xC62
IPSBAR + 0xD41
IPSBAR + 0xD42
IPSBAR + 0xD43
IPSBAR + 0xD44
IPSBAR + 0xD45
IPSBAR + 0xD46
IPSBAR + 0xD47
IPSBAR + 0xD48
IPSBAR + 0xD49
IPSBAR + 0xD4A
IPSBAR + 0xD4B
IPSBAR + 0xD4C
IPSBAR + 0xD4D
IPSBAR + 0xD4E
IPSBAR + 0xD4F
IPSBAR + 0xD51
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Size
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
Mnemonic
ICR126
ICR127
ICR128
ICR129
ICR130
ICR131
ICR132
ICR133
ICR134
ICR118
ICR119
ICR120
ICR121
ICR122
ICR123
ICR124
ICR125
ICR109
ICR110
ICR111
ICR112
ICR113
ICR114
ICR115
ICR117
ICR101
ICR102
ICR103
ICR104
ICR105
ICR106
ICR107
ICR108
A-11
Register Memory Map Quick Reference
A-12
Address
IPSBAR + 0xD73
IPSBAR + 0xD74
IPSBAR + 0xD75
IPSBAR + 0xD76
IPSBAR + 0xD77
IPSBAR + 0xD78
IPSBAR + 0xD79
IPSBAR + 0xD7A
IPSBAR + 0xD7B
IPSBAR + 0xD7C
IPSBAR + 0xD7D
IPSBAR + 0xD7E
IPSBAR + 0xDE0
IPSBAR + 0xDE4
IPSBAR + 0xDE8
IPSBAR + 0xDEC
IPSBAR + 0xDF1
IPSBAR + 0xD63
IPSBAR + 0xD64
IPSBAR + 0xD65
IPSBAR + 0xD66
IPSBAR + 0xD67
IPSBAR + 0xD68
IPSBAR + 0xD69
IPSBAR + 0xD6A
IPSBAR + 0xD6B
IPSBAR + 0xD6C
IPSBAR + 0xD6D
IPSBAR + 0xD6E
IPSBAR + 0xD6F
IPSBAR + 0xD71
IPSBAR + 0xD71
IPSBAR + 0xD72
Table A-3. Register Memory Map (continued)
Name
Interrupt Control Register 1-35
Interrupt Control Register 1-36
Interrupt Control Register 1-37
Interrupt Control Register 1-38
Interrupt Control Register 1-39
Interrupt Control Register 1-40
Interrupt Control Register 1-41
Interrupt Control Register 1-42
Interrupt Control Register 1-43
Interrupt Control Register 1-44
Interrupt Control Register 1-45
Interrupt Control Register 1-46
Interrupt Control Register 1-47
Interrupt Control Register 1-48
Interrupt Control Register 1-49
Interrupt Control Register 1-50
Interrupt Control Register 1-51
Interrupt Control Register 1-52
Interrupt Control Register 1-53
Interrupt Control Register 1-54
Interrupt Control Register 1-55
Interrupt Control Register 1-56
Interrupt Control Register 1-57
Interrupt Control Register 1-58
Interrupt Control Register 1-59
Interrupt Control Register 1-60
Interrupt Control Register 1-61
Interrupt Control Register 1-62
Software Interrupt Acknowledge Register 1
Level 1 Interrupt Acknowledge Register 1
Level 2 Interrupt Acknowledge Register 1
Level 3 Interrupt Acknowledge Register 1
Level 4 Interrupt Acknowledge Register 1
Mnemonic
ICR159
ICR160
ICR161
ICR162
SWACKR1
L1IACKR1
L2IACKR1
L3IACKR1
L4IACKR0
ICR151
ICR152
ICR153
ICR154
ICR155
ICR156
ICR157
ICR158
ICR143
ICR144
ICR145
ICR146
ICR147
ICR148
ICR149
ICR150
ICR135
ICR136
ICR137
ICR138
ICR139
ICR140
ICR141
ICR142
Size
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Address
IPSBAR + 0xDF4
IPSBAR + 0xDF8
IPSBAR + 0xDFC
IPSBAR + 0xFE4
IPSBAR + 0xFE8
IPSBAR + 0xFEC
IPSBAR + 0xFF0
IPSBAR + 0xFF4
IPSBAR + 0xFF8
IPSBAR + 0xFFC
IPSBAR + 0x10_0000
IPSBAR + 0x10_0001
IPSBAR + 0x10_0002
IPSBAR + 0x10_0003
IPSBAR + 0x10_0004
IPSBAR + 0x10_0005
IPSBAR + 0x10_0006
IPSBAR + 0x10_0007
IPSBAR + 0x10_0008
IPSBAR + 0x10_0009
IPSBAR + 0x10_000A
IPSBAR + 0x10_000B
IPSBAR + 0x10_000C
IPSBAR + 0x10_000D
IPSBAR + 0x10_000E
IPSBAR + 0x10_000F
IPSBAR + 0x10_0010
IPSBAR + 0x10_0011
IPSBAR + 0x10_0012
IPSBAR + 0x10_0013
Register Memory Map Quick Reference
Table A-3. Register Memory Map (continued)
Name
Level 5 Interrupt Acknowledge Register 1
Level 6 Interrupt Acknowledge Register 1
Level 7 Interrupt Acknowledge Register 1
Global Interrupt Acknowledge Cycle Registers
Global Level 1 Interrupt Acknowledge Register
Global Level 2 Interrupt Acknowledge Register
Global Level 3 Interrupt Acknowledge Register
Global Level 4 Interrupt Acknowledge Register
Global Level 5 Interrupt Acknowledge Register
Global Level 6 Interrupt Acknowledge Register
Global Level 7 Interrupt Acknowledge Register
GPIO Registers
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Port NQ Out Data Register
Reserved
Port AN Output Data Register
Port AS Output Data Register
Port QS Output Data Register
Reserved
Port TA Output Data Register
Port TC Output Data Register
Port TD Output Data Register
Port UA Output Data Register
Port UB Output Data Register
Port UC Output Data Register
Mnemonic
L5IACKR1
L6IACKR1
L7IACKR1
GL1IACKR
GL2IACKR
GL3IACKR
GL4IACKR
GL5IACKR
GL6IACKR
GL7IACKR
PORTNQ
—
PORTAN
PORTAS
PORTQS
—
PORTTA
PORTTC
—
—
—
—
—
—
—
—
PORTTD
PORTUA
PORTUB
PORTUC
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
Size
8
8
8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
A-13
Register Memory Map Quick Reference
A-14
Address
IPSBAR + 0x10_0014
IPSBAR + 0x10_0015
IPSBAR + 0x10_0016
IPSBAR + 0x10_0017
IPSBAR + 0x10_0018
IPSBAR + 0x10_0019
IPSBAR + 0x10_001A
IPSBAR + 0x10_001B
IPSBAR + 0x10_001C
IPSBAR + 0x10_001D
IPSBAR + 0x10_001E
IPSBAR + 0x10_001F
IPSBAR + 0x10_0020
IPSBAR + 0x10_0021
IPSBAR + 0x10_0022
IPSBAR + 0x10_0023
IPSBAR + 0x10_0024
IPSBAR + 0x10_0025
IPSBAR + 0x10_0026
IPSBAR + 0x10_0027
IPSBAR + 0x10_0028
IPSBAR + 0x10_0029
IPSBAR + 0x10_002A
IPSBAR + 0x10_002B
IPSBAR + 0x10_002C
IPSBAR + 0x10_002D
IPSBAR + 0x10_002E
IPSBAR + 0x10_002F
IPSBAR + 0x10_0030
IPSBAR + 0x10_0031
IPSBAR + 0x10_0032
IPSBAR + 0x10_0033
IPSBAR + 0x10_0034
Table A-3. Register Memory Map (continued)
Name
Port DD Output Data Register
Port LD Output Data Register
Port GP Output Data Register
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Port NQ Data Direction Register
Reserved
Port AN Data Direction Register
Port AS Data Direction Register
Port QS Data Direction Register
Reserved
Port TA Data Direction Register
Port TC Data Direction Register
Port TD Data Direction Register
Port UA Data Direction Register
Port UB Data Direction Register
Port UC Data Direction Register
Port DD Data Direction Register
Port LD Data Direction Register
Port GPData Direction Register
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Mnemonic
DDRDD
DDRLD
DDRGP
—
—
—
—
—
—
DDRQS
—
DDRTA
DDRTC
DDRTD
DDRUA
DDRUB
DDRUC
—
—
—
—
DDRNQ
—
DDRAN
DDRAS
PORTDD
PORTLD
PORTGP
—
—
—
—
—
Size
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Address
IPSBAR + 0x10_0035
IPSBAR + 0x10_0036
IPSBAR + 0x10_0037
IPSBAR + 0x10_0038
IPSBAR + 0x10_0039
IPSBAR + 0x10_003A
IPSBAR + 0x10_003B
IPSBAR + 0x10_003C
IPSBAR + 0x10_003D
IPSBAR + 0x10_003E
IPSBAR + 0x10_003F
IPSBAR + 0x10_0040
IPSBAR + 0x10_0041
IPSBAR + 0x10_0042
IPSBAR + 0x10_0043
IPSBAR + 0x10_0044
IPSBAR + 0x10_0045
IPSBAR + 0x10_0046
IPSBAR + 0x10_0047
IPSBAR + 0x10_0048
IPSBAR + 0x10_0049
IPSBAR + 0x10_004A
IPSBAR + 0x10_004B
IPSBAR + 0x10_004C
IPSBAR + 0x10_004D
Table A-3. Register Memory Map (continued)
Name
Reserved
Reserved
Reserved
Port NQ Pin Data/Set Data Register
Reserved
Port AN Pin Data/Set Data Register
Port AS Pin Data/Set Data Register
Port QS Pin Data/Set Data Register
Reserved
Port TA Pin Data/Set Data Register
Port TC Pin Data/Set Data Register
Port TD Pin Data/Set Data Register
Port UA Pin Data/Set Data Register
Port UB Pin Data/Set Data Register
Port UC Pin Data/Set Data Register
Port DD Pin Data/Set Data Register
Port LD Pin Data/Set Data Register
Port GP Pin Data/Set Data Register
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Register Memory Map Quick Reference
Mnemonic
—
—
—
PORTNQ/
SETNQ
—
PORTDDP/
SETAN
PORTELP/
SETAS
PORTQSP/
SETQS
—
PORTSDP/
SETTA
PORTTCP/
SETTC
PORTTDP/
SETTD
PORTUAP/
SETUA
PORTUAP/
SETUB
—
—
—
—
—
—
—
PORTUAP/
SETUC
PORTDDP/
SETDD
PORTLDP/
SETLD
PORTGPP/
SETGP
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Size
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
A-15
Register Memory Map Quick Reference
A-16
Address
IPSBAR + 0x10_004E
IPSBAR + 0x10_004F
IPSBAR + 0x10_0050
IPSBAR + 0x10_0051
IPSBAR + 0x10_0052
IPSBAR + 0x10_0053
IPSBAR + 0x10_0054
IPSBAR + 0x10_0055
IPSBAR + 0x10_0056
IPSBAR + 0x10_0057
IPSBAR + 0x10_0058
IPSBAR + 0x10_0059
IPSBAR + 0x10_005A
IPSBAR + 0x10_005B
IPSBAR + 0x10_005C
IPSBAR + 0x10_005D
IPSBAR + 0x10_005E
IPSBAR + 0x10_005F
IPSBAR + 0x10_0060
IPSBAR + 0x10_0061
IPSBAR + 0x10_0062
IPSBAR + 0x10_0063
IPSBAR + 0x10_0064
IPSBAR + 0x10_0065
IPSBAR + 0x10_0066
IPSBAR + 0x10_0067
IPSBAR + 0x10_0068
IPSBAR + 0x10_006A
IPSBAR + 0x10_006B
IPSBAR + 0x10_006C
IPSBAR + 0x10_006E
IPSBAR + 0x10_006F
IPSBAR + 0x10_0070
Table A-3. Register Memory Map (continued)
Name
Reserved
Reserved
Port NQ Clear Output Data Register
Reserved
Port AN Clear Output Data Register
Port AS Clear Output Data Register
Port QS Clear Output Data Register
Reserved
Port TA Clear Output Data Register
Port TC Clear Output Data Register
Port TD Clear Output Data Register
Port UA Clear Output Data Register
Port UB Clear Output Data Register
Port UC Clear Output Data Register
Port DD Clear Output Data Register
Port LD Clear Output Data Register
Port GP Clear Output Data Register
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Port NQ Pin Assignment Register
Port AN Pin Assignment Register
Port AS Pin Assignment Register
Port QS Pin Assignment Register
Port TA Pin Assignment Register
Port TC Pin Assignment Register
Port TD Pin Assignment Register
Mnemonic
—
—
PNQPAR
PANPAR
PASPAR
PQSPAR
PTAPAR
PTCPAR
PTDPAR
CLRGP
—
—
—
—
—
—
—
CLRTA
CLRTC
CLRTD
CLRUA
CLRUB
CLRUC
CLRDD
CLRLD
—
—
CLRNQ
—
CLRAN
CLRAS
CLRQS
—
Size
8
16
8
8
8
16
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Register Memory Map Quick Reference
Table A-3. Register Memory Map (continued)
Address
IPSBAR + 0x10_0071
IPSBAR + 0x10_0072
IPSBAR + 0x10_0073
IPSBAR + 0x10_0074
IPSBAR + 0x10_0075
IPSBAR + 0x10_0076
IPSBAR + 0x10_0077
IPSBAR + 0x10_0078
IPSBAR + 0x10_007A
IPSBAR + 0x10_007C
Name
Port UA Pin Assignment Register
Port UB Pin Assignment Register
Port UC Pin Assignment Register
Port DD Pin Assignment Register
Port LD Pin Assignment Register
Port GP Pin Assignment Register
Reserved
Port Wired OR Control Register
Port Drive Strength Register [47:32]
Port Drive Strength Register [31;0]
Mnemonic
Reset Control, Chip Configuration, and Power Management Registers
PUAPAR
PUBPAR
PUCPAR
PDDPAR
PLDPAR
PGPPAR
—
PWOR
PDSR
PDSR
IPSBAR + 0x11_0000
IPSBAR + 0x11_0001
IPSBAR + 0x11_0004
IPSBAR + 0x11_0007
IPSBAR + 0x11_0008
IPSBAR + 0x11_000A
IPSBAR + 0x11_000C
IPSBAR + 0x12_0000
IPSBAR + 0x12_0002
IPSBAR + 0x12_0006
IPSBAR + 0x12_0008
IPSBAR + 0x12_000C
IPSBAR + 0x13_0000
IPSBAR + 0x13_0002
IPSBAR + 0x13_0003
IPSBAR + 0x13_0004
IPSBAR + 0x13_0005
IPSBAR + 0x13_0006
IPSBAR + 0x14_0000
Reset Control Register
Reset Status Register
Chip Configuration Register
Low-Power Control Register
Reset Configuration Register
Chip Identification Register
Real Time Clock Divide Register
Clock Module Registers
Synthesizer Control Register
Synthesizer Status Register
Low Power Divider Register
Clock Control High Register
Real Time Clock Divide Register
Edge Port Registers
EPORT0 Pin Assignment Register
EPORT0 Data Direction Register
EPORT0 Interrupt Enable Register
EPORT0 Data Register
EPORT0 Pin Data Register
EPORT0 Flag Register
EPORT1 Pin Assignment Register
RCR
RSR
CCR
LPCR
RCON
CIR
RTCDF
SYNCR
SYNSR
LPDR
CCHR
RTCDR
EPPAR0
EPDDR0
EPIER0
EPDR0
EPPDR0
EPFR0
EPPAR1
Size
16
32
8
16
8
32
8
8
16
32
8
8
16
16
8
8
8
16
16
32
16
8
8
8
16
16
16
8
32
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
A-17
Register Memory Map Quick Reference
Address
IPSBAR + 0x14_0002
IPSBAR + 0x14_0003
IPSBAR + 0x14_0004
IPSBAR + 0x14_0005
IPSBAR + 0x14_0006
IPSBAR + 0x15_0000
IPSBAR + 0x15_0002
IPSBAR + 0x15_0004
IPSBAR + 0x16_0000
IPSBAR + 0x16_0002
IPSBAR + 0x16_0004
IPSBAR + 0x19_0000
IPSBAR + 0x19_0006
IPSBAR + 0x19_0007
IPSBAR + 0x19_0008
IPSBAR + 0x19_0009
IPSBAR + 0x19_000A
IPSBAR + 0x19_000C
IPSBAR + 0x19_000E
IPSBAR + 0x19_0010
IPSBAR + 0x19_0012
IPSBAR + 0x19_0200–
0x19_027E
IPSBAR + 0x19_0280–
0x19_02FE
IPSBAR + 0x19_0300–
0x19_037E
IPSBAR + 0x19_0380–
0x19_03FE
Table A-3. Register Memory Map (continued)
Name
EPORT1 Data Direction Register
EPORT1 Interrupt Enable Register
EPORT1 Data Register
EPORT1 Pin Data Register
EPORT1 Flag Register
Programmable Interrupt Timer 0 Registers
PIT Control and Status Register 0
PIT Modulus Register 0
PIT Count Register 0
Programmable Interrupt Timer 1 Registers
PIT Control and Status Register 1
PIT Modulus Register 1
PIT Count Register 1
ADC Registers
ADC Module Configuration Register
Port QA Data Register
Port QB Data Register
Port QA Data Direction Register
Port QB Data Direction Register
ADC Control Register 0
ADC Control Register 1
ADC Control Register 2
ADC Status Register 0
ADC Status Register 1
Conversion Command Word Table
Mnemonic
EPDDR1
EPIER1
EPDR1
EPPDR1
EPFR1
PCSR 0
PMR 0
PCNTR 0
PCSR 1
PMR 1
PCNTR 1
Right Justified, Unsigned Result Register
Left Justified, Signed Result Register
Left Justified, Unsigned Result Register
ADCMCR
PORTQA
PORTQB
DDRQA
DDRQB
ACR0
ACR1
ACR2
ASR0
ASR1
CCW0–
CCW63
RJURR0–
RJURR63
LJSRR0–
LJSRR63
LJURR0–
LJURR63
General Purpose Timer A Registers
16
16
16
16
16
16
8
16
16
16
16
8
8
8
16
16
64x16
64x16
64x16
64x16
Size
8
8
8
8
8
A-18
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Address
IPSBAR + 0x1A_0000
IPSABAR + 0x1A_0001
IPSBAR + 0x1A_0002
IPSBAR + 0x1A_0003
IPSBAR + 0x1A_0004
IPSBAR + 0x1A_0006
IPSBAR + 0x1A_0008
IPSBAR + 0x1A_0009
IPSBAR + 0x1A_000B
IPSBAR + 0x1A_000C
IPSBAR + 0x1A_000D
IPSBAR + 0x1A_000E
IPSBAR + 0x1A_000F
IPSBAR + 0x1A_0010
IPSBAR + 0x1A_0012
IPSBAR + 0x1A_0014
IPSBAR + 0x1A_0016
IPSBAR + 0x1A_0018
IPSBAR + 0x1A_0019
IPSBAR + 0x1A_001a
IPSBAR + 0x1A_001D
IPSBAR + 0x1A_001E
IPSBAR + 0x1B_0000
IPSBAR + 0x1B_0001
IPSBAR + 0x1B_0002
IPSBAR + 0x1B_0003
IPSBAR + 0x1B_0004
IPSBAR + 0x1B_0005
IPSBAR + 0x1B_0008
IPSBAR + 0x1B_0009
IPSBAR + 0x1B_000C
Register Memory Map Quick Reference
Table A-3. Register Memory Map (continued)
Name
GPTA IC/OC Select Register
GPTA Compare Force Register
GPTA Output Compare 3 Mask Register
GPTA Output Compare 3 Data Register
GPTA Counter Register
GPTA System Control Register 1
GPTA Toggle-on-Overflow Register
GPTA Control Register 1
GPTA Control Register 2
GPTA Interrupt Enable Register
GPTA System Control Register 2
GPTA Flag Register 1
GPTA Flag Register 2
GPTA Channel 0 Register
GPTA Channel 1 Register
GPTA Channel 2 Register
GPTA Channel 3 Register
Pulse Accumulator Control Register
Pulse Accumulator Flag Register
Pulse Accumulator Counter Register
GPTA Port Data Register
GPTA Port Data Direction Register
Pulse Width Modulator
PWM Enable Register
PWM Polarity Register
PWM Clock Select Register
PWM Prescale Clock Select Register
PWM Center Align Enable Register
PWM Control Register
PWM Scale A Register
PWM Scale B Register
PWM channel Counter Register 0
Mnemonic
GPTAIOS
GPTACFORC
GPTAOC3M
GPTAOC3D
GPTACNT
GPTASCR1
GPTATOV
GPTACTL1
GPTACTL2
GPTAIE
GPTASCR2
GPTAFLG1
GPTAFLG2
GPTAC0
GPTAC1
GPTAC2
GPTAC3
GPTAPACTL
GPTPAFLG
GPTAPACNT
GPTAPORT
GPTADDR
PWME
PWMPOL
PWMCLK
PWMPRCLK
PWMCAE
PWMCTL
PWMSCLA
PWMSCLB
PWMCNT0
8
8
8
8
8
8
8
8
8
Size
8
16
16
16
8
8
8
8
16
8
8
8
8
8
8
8
16
8
8
8
8
8
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
A-19
Register Memory Map Quick Reference
Address
IPSBAR + 0x1B_000D
IPSBAR + 0x1B_000E
IPSBAR + 0x1B_000F
IPSBAR + 0x1B_0010
IPSBAR + 0x1B_0011
IPSBAR + 0x1B_0012
IPSBAR + 0x1B_0013
IPSBAR + 0x1B_0014
IPSBAR + 0x1B_0015
IPSBAR + 0x1B_0016
IPSBAR + 0x1B_0017
IPSBAR + 0x1B_0018
IPSBAR + 0x1B_0019
IPSBAR + 0x1B_001A
IPSBAR + 0x1B_001B
IPSBAR + 0x1C_0000
IPSBAR + 0x1C_0004
IPSBAR + 0x1C_0008
IPSBAR + 0x1C_000C
IPSBAR + 0x1C_0010
IPSBAR + 0x1C_0014
IPSBAR + 0x1C_0018
IPSBAR + 0x1C_001C
IPSBAR + 0x1C_0020
IPSBAR + 0x1C_0024
IPSBAR + 0x1C_0028
IPSBAR + 0x1C_002C
IPSBAR + 0x1C_0030
IPSBAR + 0x1C_0080
IPSBAR + 0x1D_0000
Table A-3. Register Memory Map (continued)
Name
PWM channel Counter Register 1
PWM channel Counter Register 2
PWM channel Counter Register 3
PWM channel Counter Register 4
PWM channel Counter Register 5
PWM channel Counter Register 6
PWM channel Counter Register 7
PWM Channel Period Register 0
PWM Channel Period Register 1
PWM Channel Period Register 2
PWM Channel Period Register 3
PWM Channel Period Register 4
PWM Channel Period Register 5
PWM Channel Period Register 6
PWM Channel Period Register 7
FlexCAN Registers
Module Configuration Register
FlexCAN Control Register
Free Running TImer
Reserved
Rx Global Mask
Rx Buffer 14 Mask
Rx Buffer 15 Mask
Error Counter Register
Error and Status
Reserved
Interrupt Mask Register
Reserved
Interrupt Flag Register
Message Buffer 0 - Message Buffer 15
Mnemonic
PWMCNT1
PWMCNT2
PWMCNT3
PWMCNT4
PWMCNT5
PWMCNT6
PWMCNT7
PWMPER0
PWMPER1
PWMPER2
PWMPER3
PWMPER4
PWMPER5
PWMPER6
PWMPER7
CANMCR
CANCTRL
TIMER
—
RXGMASK
RX14MASK
RX15MASK
ERRCNT
ERRSTAT
—
IMASK
—
IFLAG
MBUFF0–
MBUFF15
32
32
32
32
32
16x16bytes
32
32
32
32
16
32
32
32
Size
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
Flash Registers
CFM Configuration Register CFMMCR 16
A-20
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
Register Memory Map Quick Reference
Table A-3. Register Memory Map (continued)
Address
IPSBAR + 0x1D_0002
IPSBAR + 0x1D_0008
IPSBAR + 0x1D_0010
IPSBAR + 0x1D_0014
IPSBAR + 0x1D_0018
IPSBAR + 0x1D_0020
IPSBAR + 0x1D_0024
Name
CFM Clock Divider Register
CFM Security Register
CFM Protection Register
CFM Supervisor Access Register
CFM Data Access Register
CFM User Status Register
CFM Command Register
Mnemonic
CFMCLKD
CFMSEC
CFMPROT
CFMSACC
CFMDACC
CFMUSTAT
CFMCMD
Size
IPSBAR + 0x1E_0000
IPSBAR + 0x1E_0001
IPSBAR + 0x1E_0002
IPSBAR + 0x1E_0003
Ethernet Physical Transceiver Registers
Ethernet Physical Transceiver Control Register 0
Ethernet Physical Transceiver Control Register 1
Ethernet Physical Transceiver Status Register
Reserved
Random Number Generator H/W Accelerator Registers
EPHYCTL0
EPHYCTL1
EPHYSR
—
IPSBAR + 0x1F_0000
IPSBAR + 0x1F_0004
Random Number Generator Control Register
Random Number Generator Status Register 1
RNGCR
RNGSR
32
32
IPSBAR + 0x1F_0008 Random Number Generator Entropy Register RNGER 32
IPSBAR + 0x1F_000C Random Number Generator Output Register RNGOUT 32-
1
UMR1 n , UMR2 n , and UCSR n should be changed only after the receiver/transmitter is issued a software reset command. That is, if channel operation is not disabled, undesirable results may occur.
8
8
8
8
8
8
32
32
8
32
32
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
A-21
Register Memory Map Quick Reference
A-22
MCF52235 ColdFire® Integrated Microcontroller Reference Manual, Rev. 0
Freescale Semiconductor
Preliminary
advertisement
Related manuals
advertisement