BF533_AudioInterface

BF533_AudioInterface - Microcomputer Systems 1 ADSPBF533...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Microcomputer Systems 1 ADSPBF533 EzKit Lite Audio Interface Lab #5 Audio Interface February 11, 2012 Veton Kpuska 2 ADSPBF533 EzKit Lite Audio Interface SPI SPORT0 February 11, 2012 Veton Kpuska 3 BF533 EZKit Lite Audio Interface EZKit Lite contains AD1836 Audio Codec The AD1836 audio codec provides The SPORT0 interface of the processor links with the stereo audio data input and output pins of the AD1836 codec. The processor is capable of transferring data to the audio codec in three channels of stereo audio output and two channels of multichannel/stereo 96 kHz input. timedivision multiplexed (TDM) or twowire interface (TWI) mode. The TDM mode can operate at a maximum of 48 kHz sample rate but allows simultaneous use of all input and output channels. The TWI mode allows the codec to operate at a 96 kHz sample rate but limits the output channels to two. When using TWI mode, the TSCLK0 and RSCLK0 pins, as well as the TFS0 and RFS0 pins of the processor, must be tied together external to the processor. This is accomplished with the SW9 DIP switch (see "Push Button Enable Switch (SW9)" on page 212 for more information). February 11, 2012 Veton Kpuska 4 Blackfin Audio Interface Usage TwoWire Interface (TWI) and Serial Peripheral Interface (SPI) Forward channel used to configure and control audio converters Reverse channel relays feedback info from converters SPORT's Used as data channel for audio data SPORT TX connects to audio DAC SPORT RX connects to audio ADC Fullduplex SPORT RX/TX connects to audio codec In some codecs (e.g., AC97), SPORT also can serve as the codec control channel February 11, 2012 Veton Kpuska 5 BF533 Interface with AD1871 Data Transfer & Data Transfer & Synch Synch Configuration & Configuration & Control Control February 11, 2012 Veton Kpuska 6 SW9 Switch Configuration Make sure that SW9 switches 5 and 6 are ON & SW12 is ON. See explanation bellow: February 11, 2012 Veton Kpuska 7 Switch Default Configuration February 11, 2012 Veton Kpuska 8 Pushbuttons and LED's February 11, 2012 Veton Kpuska 9 Push Button Enable Switch (SW9 & SW12) SW9 Positions 1 through 4: Positions 5 and 6: Disconnects the drivers associated with the push buttons from the PF pins of the processor. Used to connect the transmit and receive frame syncs and clocks of SPORT0. This is important when the AD1836 audio/video decoder and the processor are communicating in I2S mode. SW12 When is set to OFF, SW12 disconnects SPORT0 from the audio codec. The default is the ON position. Table 26 in the next slide shows which PF is driven when the switch is in the default (ON) position. February 11, 2012 Veton Kpuska 10 Push Button Enable Switch (SW9) Table 26 shows which PF is driven when the switch is in the default (ON) position February 11, 2012 Veton Kpuska 11 SPORT0 interface to AD1836 audio codec The SPORT0 connects to the AD1836 audio codec and the expansion interface. The AD1836 codec uses both the primary and secondary data transmit and receive pins to input and output data from the audio inputs and outputs. The SPORT1 connects to the SPORT connector (P3) and the expansion interface. The pinout of the SPORT connector and the expansion interface connectors can be found in "ADSPBF533 EZ KIT Lite Schematic" on page B1. Veton Kpuska 12 February 11, 2012 The pin layout of the SPORT connector & expansion interface connectors SW12 February 11, 2012 Veton Kpuska 13 Audio Interface Configuration The AD1836 audio codec's internal configuration registers are configured using the Serial Port Interface SPI port of the processor ( see Slide #6). The processor's PF4 programmable flag pin is used as the select for this device. For information on how to configure the multi channel codec, go to www.analog.com/UploadedFiles/Datasheets/34474 0003AD1836_prc.pdf. February 11, 2012 Veton Kpuska 14 Audio Interface and Programmable Flags The processor has 15 programmable flag pins (PF's). The pins are multifunctional and depend on the processor setup. Table 21 in the ADSPBF533 EZKIT Lite Evaluation System Manual provides the summary of the programmable flag pins used on the EZKit Lite. February 11, 2012 Veton Kpuska 15 Programmable Flag Connections February 11, 2012 Veton Kpuska 16 Audio Interface Configuration (cont.) The generalpurpose IO pin PA0 of flash A is a source for the AD1836 codec reset. See "Flash GeneralPurpose IO" on page 112 for more information about the pin of ADSPBF533 EZKIT Lite Evaluation System Manual. Also information given in the next slide. February 11, 2012 Veton Kpuska 17 AD1836 codec reset via PA0 February 11, 2012 Veton Kpuska 18 Flash GeneralPurpose IO Generalpurpose IO signals are controlled by means of setting appropriate registers of the flash A or flash B. These registers are mapped into the processor's address space, as shown in the Table below (ADSPBF533 EZKIT Lite Evaluation System Manual Flash Memory) February 11, 2012 Veton Kpuska 19 Flash GeneralPurpose IO Flash device IO pins are arranged as 8bit ports labeled A through G. There is a set of 8bit registers associated with each port. These registers are The Direction register controls IO pins direction. This is a 8bit readwrite register. Direction, Data In, and Data Out. Note that the Direction and Data Out registers are cleared to all zeros at powerup or hardware reset. When a bit is 0, a corresponding pin functions as an input. When the bit is 1, a corresponding pin is an output. The Data In register allows reading the status of port's pins. This is a 8bit read only register. The Data Out register allows clearing an output pin to 0 or setting it to 1. This is a 8bit readwrite register. The ADSPBF533 EZKIT Lite board employs only flash A and flash B ports A and B. Table 15 and Table 16 provide configuration register addresses for flash A and flash B, respectively (only ports A and B are listed in the next slide). The following bits connect to the expansion board connector. Flash A: port A bits 7 and 6, as well as port B bits 7 and 6 Flash B: port A bits 70 February 11, 2012 Veton Kpuska 20 Flash A & B Configuration Registers for Ports A and B. Table 15. Flash A Configuration Registers for Ports A and B Register Name Data In (readonly) Data Out (readwrite) Direction (readwrite) Port A Address 0x2027 0000 0x2027 0004 0x2027 0006 Port B Address 0x2027 0001 0x2027 0005 0x2027 0007 Table 16. Flash B Configuration Registers for Ports A and B Register Name Data In (readonly) Data Out (readwrite) Direction (readwrite) Port A Address 0x202E 0000 0x202E 0004 0x202E 0006 Port B Address 0x202E 0001 0x202E 0005 0x202E 0007 February 11, 2012 Veton Kpuska 21 IO Assignments for Port A and B Table 17 Flash A Port A Controls. User IO Not defined Not defined PPI clock select bit 1 PPI clock select bit 0 Video decoder reset Video encoder reset Reserved Codec reset Bit Number 7 6 5 4 3 2 1 0 Bit Value Any Any 00=local OSC (27MHz) 01=video decoder pixel clock 1x=expansion board PPI clock 0=reset ON; 1=reset OFF 0=reset ON; 1=reset OFF Any 0=reset ON; 1=reset OFF February 11, 2012 Veton Kpuska 22 TalkThrough Analysis of Lab Exercise main.c //--------------------------------------------------------------------------// // // // Name: Talkthrough for the ADSP-BF533 EZ-KIT Lite // // // //--------------------------------------------------------------------------// // // // (C) Copyright 2003 - Analog Devices, Inc. All rights reserved. // // // // Project Name: BF533 C Talkthrough TDM // // // // Date Modified: 04/03/03 HD Rev 1.0 // // // // Software: VisualDSP++3.1 // // // // Hardware: ADSP-BF533 EZ-KIT Board // // // // Connections: Connect an input source (such as a radio) to the Audio // // input jack and an output source (such as headphones) to // // the Audio output jack // // // // Purpose: This program sets up the SPI port on the ADSP-BF533 to // // configure the AD1836 codec. The SPI port is disabled // // after initialization. The data to/from the codec are // // transfered over SPORT0 in TDM mode // // // //--------------------------------------------------------------------------// #include #include #include #include "Talkthrough.h" "sysreg.h" "ccblkfn.h" <btc.h> February 11, 2012 Veton Kpuska 24 Talkthrough.h #ifndef __Talkthrough_DEFINED #define __Talkthrough_DEFINED //--------------------------------------------------------------------------// // Header files // //--------------------------------------------------------------------------// #include <sys\exception.h> #include <cdefBF533.h> #include <ccblkfn.h> #include <sysreg.h> //--------------------------------------------------------------------------// // Symbolic constants // //--------------------------------------------------------------------------// // addresses for Port B in Flash A #define pFlashA_PortA_Dir (volatile unsigned char *)0x20270006 #define pFlashA_PortA_Data (volatile unsigned char *)0x20270004 // names for codec registers, used for iCodec1836TxRegs #define DAC_CONTROL_1 0x0000 #define DAC_CONTROL_2 0x1000 #define DAC_VOLUME_0 0x2000 #define DAC_VOLUME_1 0x3000 #define DAC_VOLUME_2 0x4000 #define DAC_VOLUME_3 0x5000 #define DAC_VOLUME_4 0x6000 #define DAC_VOLUME_5 0x7000 #define ADC_0_PEAK_LEVEL 0x8000 #define ADC_1_PEAK_LEVEL 0x9000 #define ADC_2_PEAK_LEVEL 0xA000 #define ADC_3_PEAK_LEVEL 0xB000 #define ADC_CONTROL_1 0xC000 #define ADC_CONTROL_2 0xD000 #define ADC_CONTROL_3 0xE000 February 11, 2012 Veton Kpuska 25 Flash A & B Configuration Registers for Ports A and B. Table 15. Flash A Configuration Registers for Ports A and B Register Name Data In (readonly) Data Out (readwrite) Direction (readwrite) Port A Address 0x2027 0000 0x2027 0004 0x2027 0006 Port B Address 0x2027 0001 0x2027 0005 0x2027 0007 Table 16. Flash B Configuration Registers for Ports A and B Register Name Data In (readonly) Data Out (readwrite) Direction (readwrite) Port A Address 0x202E 0000 0x202E 0004 0x202E 0006 Port B Address 0x202E 0001 0x202E 0005 0x202E 0007 February 11, 2012 Veton Kpuska 26 DAC Control Register 1 & 2 From: AD1836A Multichannel Codec Manual #define DAC_CONTROL_1 0x0000 #define DAC_CONTROL_1 0x0000 #define DAC_CONTROL_2 0x1000 #define DAC_CONTROL_2 0x1000 February 11, 2012 Veton Kpuska 27 DAC Volume Control Registers #define DAC_VOLUME_0 #define DAC_VOLUME_0 #define DAC_VOLUME_1 #define DAC_VOLUME_1 #define DAC_VOLUME_2 #define DAC_VOLUME_2 #define DAC_VOLUME_3 #define DAC_VOLUME_3 #define DAC_VOLUME_4 #define DAC_VOLUME_4 #define DAC_VOLUME_5 #define DAC_VOLUME_5 0x2000 0x2000 0x3000 0x3000 0x4000 0x4000 0x5000 0x5000 0x6000 0x6000 0x7000 0x7000 February 11, 2012 Veton Kpuska 28 ADC Peak Level Registers #define ADC_0_PEAK_LEVEL #define ADC_0_PEAK_LEVEL #define ADC_1_PEAK_LEVEL #define ADC_1_PEAK_LEVEL #define ADC_2_PEAK_LEVEL #define ADC_2_PEAK_LEVEL #define ADC_3_PEAK_LEVEL #define ADC_3_PEAK_LEVEL 0x8000 0x8000 0x9000 0x9000 0xA000 0xA000 0xB000 0xB000 February 11, 2012 Veton Kpuska 29 ADC Control Register 1 #define ADC_CONTROL_1 #define ADC_CONTROL_1 #define ADC_CONTROL_2 #define ADC_CONTROL_2 #define ADC_CONTROL_3 #define ADC_CONTROL_3 0xC000 0xC000 0xD000 0xD000 0xE000 0xE000 February 11, 2012 Veton Kpuska 30 ADC Control Register 2 #define ADC_CONTROL_1 #define ADC_CONTROL_1 #define ADC_CONTROL_2 #define ADC_CONTROL_2 #define ADC_CONTROL_3 #define ADC_CONTROL_3 0xC000 0xC000 0xD000 0xD000 0xE000 0xE000 February 11, 2012 Veton Kpuska 31 ADC Control Register 3 #define ADC_CONTROL_1 #define ADC_CONTROL_1 #define ADC_CONTROL_2 #define ADC_CONTROL_2 #define ADC_CONTROL_3 #define ADC_CONTROL_3 0xC000 0xC000 0xD000 0xD000 0xE000 0xE000 February 11, 2012 Veton Kpuska 32 Talkthrough.h (cont. 1) // names for slots in ad1836 audio frame #define INTERNAL_ADC_L0 #define INTERNAL_ADC_L1 #define INTERNAL_ADC_R0 #define INTERNAL_ADC_R1 #define INTERNAL_DAC_L0 #define INTERNAL_DAC_L1 #define INTERNAL_DAC_L2 #define INTERNAL_DAC_R0 #define INTERNAL_DAC_R1 #define INTERNAL_DAC_R2 // size of array iCodec1836TxRegs and iCodec1836RxRegs #define CODEC_1836_REGS_LENGTH 11 // SPI transfer mode #define TIMOD_DMA_TX 0x0003 // SPORT0 word length #define SLEN_32 // DMA flow mode #define FLOW_1 0x001f 0x1000 0 1 4 5 0 1 2 4 5 6 February 11, 2012 Veton Kpuska 33 Talkthrough.h (cont. 2) //// // Global variables //// extern int iChannel0LeftIn; extern int iChannel0RightIn; extern int iChannel0LeftOut; extern int iChannel0RightOut; extern int iChannel1LeftIn; extern int iChannel1RightIn; extern int iChannel1LeftOut; extern int iChannel1RightOut; extern volatile short sCodec1836TxRegs; extern volatile int iRxBuffer1; extern volatile int iTxBuffer1; //// // Prototypes //// // in file Initialisation.c void Init_EBIU(void); void Init_Flash(void); void Init1836(void); void Init_Sport0(void); void Init_DMA(void); void Init_Sport_Interrupts(void); void Enable_DMA_Sport0(void); // in file Process_data.c void Process_Data(void); // in file ISRs.c EX_INTERRUPT_HANDLER(Sport0_RX_ISR); #endif //__Talkthrough_DEFINED // // February 11, 2012 Veton Kpuska 34 main.c #include "Talkthrough.h" #include "sysreg.h" #include "ccblkfn.h" #include <btc.h> //--------------------------------------------------------------------------// // Variables // // // // Description: The variables iChannelxLeftIn and iChannelxRightIn contain // // the data coming from the codec AD1836. The (processed) // // playback data are written into the variables // // iChannelxLeftOut and iChannelxRightOut respectively, which // // are then sent back to the codec in the SPORT0 ISR. // // The values in the array iCodec1836TxRegs can be modified to // // set up the codec in different configurations according to // // the AD1885 data sheet. // //--------------------------------------------------------------------------// // left input data from ad1836 int iChannel0LeftIn, iChannel1LeftIn; // right input data from ad1836 int iChannel0RightIn, iChannel1RightIn; // left ouput data for ad1836 int iChannel0LeftOut, iChannel1LeftOut; // right ouput data for ad1836 int iChannel0RightOut, iChannel1RightOut; February 11, 2012 Veton Kpuska 35 main.c (cont 1) // array for registers to configure the ad1836 names are defined in "Talkthrough.h" volatile short sCodec1836TxRegs[CODEC_1836_REGS_LENGTH] = { DAC_CONTROL_1 | 0x000, DAC_CONTROL_2 | 0x000, DAC_VOLUME_0| 0x3ff, DAC_VOLUME_1| 0x3ff, DAC_VOLUME_2| 0x3ff, DAC_VOLUME_3| 0x3ff, DAC_VOLUME_4| 0x3ff, DAC_VOLUME_5| 0x3ff, ADC_CONTROL_1 | 0x000, ADC_CONTROL_2 | 0x180, ADC_CONTROL_3 | 0x000 }; #define DAC_CONTROL_1 #define DAC_VOLUME_0 DAC_CONTROL_1 DAC_VOLUME_0 #define DAC_CONTROL_2 DAC_VOLUME_1 DAC_CONTROL_2 #define DAC_VOLUME_1 #define DAC_VOLUME_2 #define DAC_VOLUME_2 #define DAC_VOLUME_3 #define DAC_VOLUME_3 #define DAC_VOLUME_4 #define DAC_VOLUME_4 #define DAC_VOLUME_5 #define DAC_VOLUME_5 #define ADC_CONTROL_1 #define ADC_CONTROL_1 #define ADC_CONTROL_2 #define ADC_CONTROL_2 #define ADC_CONTROL_3 #define ADC_CONTROL_3 0x0000 || 0x000 == 0x23ff 0x2000 0x0000 0x3ff 0x000 0x0000 0x0000 0x2000 0x3ff 0x23ff 0x1000 | 0x3ff = 0x33ff 0x3000 | 0x000 = 0x1000 0x1000 0x000 0x1000 0x3000 0x3ff 0x33ff 0x4000 || 0x3ff == 0x43ff 0x4000 0x3ff 0x43ff 0x5000 || 0x3ff == 0x53ff 0x5000 0x3ff 0x53ff 0x6000 || 0x3ff == 0x63ff 0x6000 0x3ff 0x63ff 0x7000 || 0x3ff == 0x73ff 0x7000 0x3ff 0x73ff 0xC000 || 0x000 == 0xC000 0xC000 0x000 0xC000 0xD000 || 0x180 == 0xD180 0xD000 0x180 0xD180 0xE000 || 0x000 == 0xE000 0xE000 0x000 0xE000 February 11, 2012 Veton Kpuska 36 ADC Master Mode February 11, 2012 Veton Kpuska 37 ADC Slave Mode February 11, 2012 Veton Kpuska 38 main.c (cont 2) // SPORT0 DMA transmit buffer volatile int iTxBuffer1[8]; // SPORT0 DMA receive buffer volatile int iRxBuffer1[8]; //////////////////// // BTC Definitions //////////////////// #define MAXBTCBUF 24000 int BTCLeft[MAXBTCBUF]; int BTCRight[MAXBTCBUF]; int BTCLeftVolume = 0; int BTCRightVolume = 0; BTC_MAP_BEGIN // Channel Name, Starting Address, Length (bytes) BTC_MAP_ENTRY("Audio Left Channel", (long)&BTCLeft, sizeof(BTCLeft)) BTC_MAP_ENTRY("Audio Right Channel", (long)&BTCRight, sizeof(BTCRight)) BTC_MAP_ENTRY("Audio Left Volume", (long)&BTCLeftVolume, sizeof(BTCLeftVolume)) BTC_MAP_ENTRY("Audio Right Volume", (long)&BTCRightVolume, sizeof(BTCRightVolume)) BTC_MAP_END February 11, 2012 Veton Kpuska 39 main.c (cont. 3) //--------------------------------------------------------------------------// // Function: main // // // // Description: After calling a few initalization routines, main() just // // waits in a loop forever. The code to process the incoming // // data can be placed in the function Process_Data() in the // // file "Process_Data.c". // //--------------------------------------------------------------------------// void main(void) { sysreg_write(reg_SYSCFG, 0x32); Init_EBIU(); Init_Flash(); Init1836(); Init_Sport0(); Init_DMA(); Init_Sport_Interrupts(); Enable_DMA_Sport0(); btc_init(); while(1) { btc_poll(); } } //Initialize System Configuration Register February 11, 2012 Veton Kpuska 40 Background Telemetry Channels (BTCs) Background telemetry channels (BTCs) enable VisualDSP++ and a processor to exchange data via the JTAG interface while the processor is executing. Before BTC, all communication between VisualDSP++ and a processor took place while the processor was in a halted state. Note: Background telemetry channels are supported only in SHARC and Blackfin emulator sessions. Veton Kpuska 41 February 11, 2012 BTC Definitions in the Program Background telemetry channels are defined on a per program (.DXE) basis. The channels are defined when a specific program is loaded onto a processor. BTC channels are defined in a program by using simple macros. The following example code shows channel definitions. #include "btc.h" BTC_MAP_BEGIN // Channel Name, Starting Address, Length (bytes) BTC_MAP_ENTRY("Audio Left Channel", (long)&BTCLeft, sizeof(BTCLeft)) BTC_MAP_ENTRY("Audio Right Channel", (long)&BTCRight, sizeof(BTCRight)) BTC_MAP_ENTRY("Audio Left Volume", (long)&BTCLeftVolume, sizeof(BTCLeftVolume)) BTC_MAP_ENTRY("Audio Right Volume", (long)&BTCRightVolume, sizeof(BTCRightVolume)) BTC_MAP_END The first step in defining channels in a program is to include the BTC macros by using the #include btc.h statement. Then each channel is defined with the macros. The definitions begin with BTC_MAP_BEGIN, which marks the beginning of the BTC map. Next, each individual channel is defined with the BTC_MAP_ENTRY macro, which takes the parameters described in Table 217 in VisualDSP++ 4.5 User's Guide. February 11, 2012 Veton Kpuska 42 BTC_MAP_ENTRY Macro February 11, 2012 Veton Kpuska 43 Initialize.c 1. 2. 3. 4. 5. 6. 7. 8. 9. EBIU setup Flash setup FIO setup ADI1836 SPORT0 DMA configuration SPORT IRS DMA Support Endless loop and processing Flag Interrupts External Bus Interface Unit (EBIU) Setup #include "Talkthrough.h" //--------------------------------------------------------------------------// // Function: Init_EBIU // // // // Description: This function initializes and enables asynchronous memory // // banks in External Bus Interface Unit so that Flash A can be // // accessed. // //--------------------------------------------------------------------------// void Init_EBIU(void) { *pEBIU_AMBCTL0 = 0x7bb07bb0; *pEBIU_AMBCTL1 = 0x7bb07bb0; *pEBIU_AMGCTL = 0x000f; } February 11, 2012 Veton Kpuska 45 External Bus Interface Unit The EBIU services requests for external memory from the Core or from a DMA channel. The priority of the requests is determined by the External Bus Controller. The address of the request determines whether the request is serviced by the EBIU SDRAM Controller or the EBIU Asynchronous Memory Controller. The EBIU is clocked by the system clock (SCLK). All synchronous memories interfaced to the processor operate at the SCLK frequency. The ratio between core frequency and SCLK frequency is programmable using a phaselocked loop (PLL) system memory mapped register (MMR). February 11, 2012 Veton Kpuska 46 EBIU Programming Model This programming model is based on system memorymapped registers used to program the EBIU. There are six control registers and one status register in the EBIU. They are: Asynchronous Memory Global Control register (EBIU_AMGCTL) Asynchronous Memory Bank Control 0 register (EBIU_AMBCTL0) Asynchronous Memory Bank Control 1 register (EBIU_AMBCTL1) SDRAM Memory Global Control register (EBIU_SDGCTL) SDRAM Memory Bank Control register (EBIU_SDBCTL) SDRAM Refresh Rate Control register (EBIU_SDRRC) SDRAM Control Status register (EBIU_SDSTAT) February 11, 2012 Veton Kpuska 47 The Asynchronous Memory Interface The asynchronous memory interface allows interfaceing to a variety of memory and peripheral types. These include SRAM, ROM, EPROM, flash memory, and FPGA/ASIC designs. Four asynchronous memory regions are supported. Each has a unique memory select associated with it, shown in Table 173. February 11, 2012 Veton Kpuska 48 Asynchronous Memory Bank Address Range February 11, 2012 Veton Kpuska 49 EBIU_AMGCTL Register The Asynchronous Memory Global Control register (EBIU_AMGCTL) configures global aspects of the controller. The EBIU_AMGCTL register should be the last control register written to when configuring the processor to access external memorymapped asynchronous devices. February 11, 2012 Veton Kpuska 50 Initialization.c //--------------------------------------------------------------------------// // Function: Init_EBIU // // // // Parameters: None // // // // Return: None // // // // Description: This function initializes and enables the asynchronous // // memory banks for the External Bus Interface Unit (EBIU), so // // that access to Flash A is possible. // //--------------------------------------------------------------------------// void Init_EBIU(void) { *pEBIU_AMBCTL0 = 0x7bb07bb0; *pEBIU_AMBCTL1 = 0x7bb07bb0; Asynchronous Memory *pEBIU_AMGCTL = 0x000f; Asynchronous Memory Global Control Register } Global Control Register February 11, 2012 Veton Kpuska 51 Asynchronous Memory Global Control Register EBIU_AMGCTL 0x000f = 0000 0000 0000 1111 0x000f = 0000 0000 0000 1111 February 11, 2012 Veton Kpuska 52 Initialization.c //--------------------------------------------------------------------------// // Function: Init_EBIU // // // // Parameters: None // // // // Return: None // // // // Description: This function initializes and enables the asynchronous // // memory banks for the External Bus Interface Unit (EBIU), so // // that access to Flash A is possible. // //--------------------------------------------------------------------------// void Init_EBIU(void) { *pEBIU_AMBCTL0 = 0x7bb07bb0; Asynchronous Memory Asynchronous Memory *pEBIU_AMBCTL1 = 0x7bb07bb0; Control Register Control Register *pEBIU_AMGCTL = 0x000f; } February 11, 2012 Veton Kpuska 53 EBIU_AMBCTL0 and EBIU_AMBCTL1 Registers The EBIU asynchronous memory controller has two Asynchronous Memory Bank Control registers: They contain bits for counters for: EBIU_AMBCTL0 and EBIU_AMBCTL1. setup, strobe, and hold time; bits to determine memory type and size; and bits to configure use of ARDY. Veton Kpuska 54 February 11, 2012 Asynchronous Memory Bank Control 0 Register 0x7bb07bb0 = 0x7bb07bb0 = 0111 1011 1011 0000 0111 1011 1011 0000 0111 1011 1011 0000 0111 1011 1011 0000 February 11, 2012 Veton Kpuska 55 External Memory Map The three SDRAM control registers must be initialized in order to use the MT48LC32M16 64 MB (32M x 16 bits) SDRAM memory of EZKit Lite: EBIU_SDGCTL: SDRAM Memory Global Control register EBIU_SDBCTL: SDRAM Memory Bank Control register EBIU_SDRRC: SDRAM Refresh Rate Control register The External Memory Map supported by BF533 is provided in the Figure 171. February 11, 2012 Veton Kpuska 56 EBIU_SDGCTL Register The SDRAM Memory Global Control register (EBIU_SDGCTL) includes all programmable parameters associated with the SDRAM access timing and configuration. Figure 1710 shows the EBIU_SDGCTL register bit definitions. February 11, 2012 Veton Kpuska 57 SDRAM Memory Global Control register: EBIU_SDGCTL February 11, 2012 Veton Kpuska 58 Initialize.c 1. 2. 3. 4. 5. 6. 7. 8. 9. EBIU setup Flash setup FIO setup ADI1836 SPORT0 DMA configuration SPORT IRS DMA Support Endless loop and processing Flag Interrupts Initialize.c //--------------------------------------------------------------------------// // Function: Init_Flash // // // // Description: This function initializes pin direction of Port A in Flash A // // to output. The AD1836_RESET on the ADSP-BF533 EZ-KIT board // // is connected to Port A. // //--------------------------------------------------------------------------// void Init_Flash(void) { *pFlashA_PortA_Dir = 0x1; } February 11, 2012 Veton Kpuska 60 Initialize.c //--------------------------------------------------------------------------// // Function: Init_Flash // // // // Description: This function initializes pin direction of Port A in Flash A // // to output. The AD1836_RESET on the ADSP-BF533 EZ-KIT board // // is connected to Port A. // //--------------------------------------------------------------------------// void Init_Flash(void) { Direction Direction 0x1 = 0001 *pFlashA_PortA_Dir = 0x1; 0x1 = 0001 Configuration } Configuration The Direction register controls IO pins direction. The Direction register controls IO pins direction. This is a 8bit readwrite register. This is a 8bit readwrite register. When a bit is 0, a corresponding pin functions as an input. When a bit is 0, a corresponding pin functions as an input. When the bit is 1, a corresponding pin is an output. When the bit is 1, a corresponding pin is an output. See Slide #13 for details or EZKit Lite HRM. See Slide #13 for details or EZKit Lite HRM. February 11, 2012 Veton Kpuska 61 Initialize.c 1. 2. 3. 4. 5. 6. 7. 8. 9. EBIU setup Flash setup FIO setup ADC1836 SPORT0 DMA configuration SPORT IRS DMA Support Endless loop and processing Flag Interrupts BF533 Interface with AD1871 Data Transfer & Data Transfer & Synch Synch Configuration & Configuration & Control Control February 11, 2012 Veton Kpuska 63 Initialize.c //// // Function: Init1836() // // // // Description: This function sets up the SPI port to configure the AD1836. // // The content of the array sCodec1836TxRegs is sent to the // // codec. // //// void Init1836(void) { int i; int j; static unsigned char ucActive_LED = 0x01; // write to Port A to reset AD1836 *pFlashA_PortA_Data = 0x00; // write to Port A to enable AD1836 *pFlashA_PortA_Data = ucActive_LED; // wait to recover from reset for (i=0; i<0xf000; i++); // Enable PF4 *pSPI_FLG = FLS4; // Set baud rate SCK = HCLK/(2*SPIBAUD) SCK = 2MHz *pSPI_BAUD = 16; // configure spi port // SPI DMA write, 16bit data, MSB first, SPI Master *pSPI_CTL = TIMOD_DMA_TX | SIZE | MSTR; Reset AD1836 ADC & Reset AD1836 ADC & DAC Codec. DAC Codec. Activate AD1836 Via PF4 Activate AD1836 Via PF4 pin pin Setting the Baud Rate of Setting the Baud Rate of the Codec the Codec Setting the DMA transfer Setting the DMA transfer February 11, 2012 Veton Kpuska 64 Serial Peripheral Interface (SPI) The processor has a Serial Peripheral Interface (SPI) port that provides an I/O interface to a wide variety of SPI compatible peripheral devices. SPI is a fourwire interface consisting of two data pins (IN & OUT), a device select pin (SPISS), and a clock (CLK) pin. SPI is a fullduplex synchronous serial interface, supporting master modes, slave modes, and multimaster environments. programmable baud rate and clock phase/polarities. The SPI compatible peripheral implementation also supports February 11, 2012 Veton Kpuska 65 Serial Peripheral Interface (SPI) Typical SPI compatible peripheral devices that can be used to interface to the SPI compatible interface include: Other CPUs or microcontrollers Codecs A/D converters D/A converters Sample rate converters SP/DIF or AES/EBU digital audio transmitters and receivers LCD displays Shift registers FPGAs with SPI emulation Veton Kpuska 66 February 11, 2012 Serial Peripheral Interface (SPI) The SPI is an industrystandard synchronous serial link that supports communication with multiple SPI compatible devices. The SPI peripheral is a synchronous, fourwire interface consisting of two data pins: one device select pin Master Out Slave In (MOSI) and Master In Slave Out (MISO) Serial Peripheral Interface Slave Select (SPISS) Input Signal, and a Serial Peripheral Interface Clock (SCK) Signal. gated clock pin With the two data pins, it allows for fullduplex operation to other SPI compatible devices. The SPI also includes programmable baud rates, clock phase, and clock polarity. February 11, 2012 Veton Kpuska 67 Serial Peripheral Interface (SPI) The SPI can operate in a multimaster environment by interfacing with several other devices, acting as either a master device or a slave device. Figure 101 provides a block diagram of the SPI. The interface is essentially a shift register that serially transmits and receives data bits, one bit at a time at the SCK rate, to and from other SPI devices. SPI data is transmitted and received at the same time through the use of a shift register. When an SPI transfer occurs, data is simultaneously transmitted (shifted serially out of the shift register) as new data is received (shifted serially into the other end of the same shift register). The SCK synchronizes the shifting and sampling of the data on the two serial data pins. February 11, 2012 Veton Kpuska 68 SPI Block Diagram February 11, 2012 Veton Kpuska 69 SPI Interface Signals 1. 2. 3. 4. 5. 6. 7. 8. 9. EBIU setup Flash setup FIO setup ADI1836 SPI SPORT0 DMA configuration SPORT IRS DMA Support Endless loop and processing Flag Interrupts Serial Peripheral Interface Clock Signal (SCK) The SCK signal is the SPI clock signal. This control signal is driven by the master and controls the rate at which data is transferred. The master may transmit data at a variety of baud rates. It is an The SCK signal cycles once for each bit transmitted. output signal if the device is configured as a master, and an input signal if the device is configured as a slave. The SCK is a gated clock that is active during data transfers only for the length of the transferred word. The number of active clock edges is equal to the number of bits driven on the data lines. Slave devices ignore the serial clock if the Serial Peripheral Slave Select Input (SPISS') is driven inactive (high). The SCK is used to shift out and shift in the data driven on the MISO and MOSI lines. The data is always shifted out on active edges of the clock and sampled on inactive edges of the clock. Clock polarity and clock phase relative to data are programmable in the SPI Control register (SPI_CTL) and define the transfer format. February 11, 2012 Veton Kpuska 71 Serial Peripheral Interface Slave Select Input Signal The SPISS' signal is the SPI Serial Peripheral Slave Select Input signal. This is an activelow signal used to enable a processor when it is configured as a slave device. This inputonly pin behaves like a chip select and is provided by the master device for the slave devices. For a master device, it can act as an error signal input in case of the multimaster environment. In multimaster mode, if the SPISS input signal of a master is asserted (driven low), and the PSSE bit in the SPI_CTL register is enabled, an error has occurred. This means that another device is also trying to be the master device. Note: The SPISS signal is the same pin as the PF0 pin. Veton Kpuska 72 February 11, 2012 Master Out Slave In (MOSI) & Master In Slave Out (MISO) Master Out Slave In (MOSI) The MOSI pin is the Master Out Slave In pin, one of the bidirectional I/O data pins. If the processor is configured as a master, the MOSI pin becomes a data transmit (output) pin, transmitting output data. If the processor is configured as a slave, the MOSI pin becomes a data receive (input) pin, receiving input data. In an SPI interconnection, the data is shifted out from the MOSI output pin of the master and shifted into the MOSI input(s) of the slave(s). The MISO pin is the Master In Slave Out pin, one of the bidirectional I/O data pins. Master In Slave Out (MISO) If the processor is configured as a master, the MISO pin becomes a data receive (input) pin, receiving input data. If the processor is configured as a slave, the MISO pin becomes a data transmit (output) pin, February 11, 2012 Veton Kpuska 73 BF533 Interface with AD1871 Data Transfer & Data Transfer & Synch Synch Configuration & Configuration & Control Control February 11, 2012 Veton Kpuska 74 ADSPBF533 as Slave SPI Device February 11, 2012 Veton Kpuska 75 Interrupt Output The SPI has two interrupt output signals: a data interrupt and an error interrupt February 11, 2012 Veton Kpuska 76 SPI Data Interrupt Output The behavior of the SPI data interrupt signal depends on the Transfer Initiation Mode bit field (TIMOD) in the SPI Control register (discussed next). In DMA mode (TIMOD = 1X), the data interrupt acts as a DMA request and is generated when the DMA FIFO is ready to be written to (TIMOD = 11) or read from (TIMOD = 10). In nonDMA mode (TIMOD = 0X), a data interrupt is generated when the SPI_TDBR is ready to written to (TIMOD = 01) or when the SPI_RDBR is ready to be read from (TIMOD = 00). February 11, 2012 Veton Kpuska 77 SPI Error Interrupt Output An SPI Error interrupt is generated in a master when a Mode Fault Error occurs, in both DMA and nonDMA modes. An error interrupt can also be generated in DMA mode when there is In nonDMA mode, the underflow and overflow conditions set the TXE and RBSY bits in the SPI_STAT register, respectively, but do not generate an error interrupt. Veton Kpuska 78 an underflow (TXE when TIMOD = 11) or an overflow (RBSY when TIMOD = 10) error condition. February 11, 2012 SPI Registers The SPI peripheral includes a number of useraccessible registers. Some of these registers are also accessible through the DMA bus. Four registers contain control and status information: SPI_BAUD, SPI_CTL, SPI_FLG, and SPI_STAT. Two registers are used for buffering receive and transmit data: SPI_RDBR and SPI_TDBR For information about DMArelated registers, see Chapter 9 of ADSPBF533 Blackfin Processor Hardware Reference, "Direct Memory Access." also discussed next. The shift register, SFDR, is internal to the SPI module and is not directly accessible. February 11, 2012 Veton Kpuska 79 SPI Registers: SPI_BAUD Register The SPI Baud Rate register (SPI_BAUD) is used to set the bit transfer rate for a master device. The serial clock frequency is determined by this formula: When configured as a slave, the value written to this register is ignored. SCK Frequency = (Peripheral clock frequency SCLK)/(2 x SPI_BAUD) SCK Clock Divide Factor SPI_BAUD = 16 & assuming SCLK = 100 MHz SCK = 100 MHz/32 = 3125 Hz. February 11, 2012 Veton Kpuska 80 SPI_BAUD Register February 11, 2012 Veton Kpuska 81 Note on SPORT Clock division and frame sync The clock division parameters are located in the To specify a clock frequency for the SPORT, use the following equation: SPORTx_TCLKDIV and SPORTx_RCLKDIV registers. f ( sclk ) SPORTx _ yCLK _ frequency = 2 * ( SPORTx _ yCLKDIV + 1) f(sclk) is the frequency of the core (~52 Mhz default on the EZKIT LITE), SPORTx_yCLK (y is T for transmit, R for receive) frequency is the specified clock frequency of the SPORT, and SPORTx_yCLKDIV is the value needed to be input into the specified transmit/receive clock division register. Where: Frame sync parameters are the number of serial clock cycles before a frame sync is generated. This value must not be less than the serial word length (SLEN), otherwise corrupt data will be received. # of serial clock cycles between frame syncs = SPORTx_yFSDIV + 1 For example, if one wanted a serial clock frequency of 375 kHz, and 32 bit words were used, then SPORTx_yCLKDIV = 52MHz/(2*375kHz) 1 = 70 1 = 69 = 0x0045 SPORTx_yFSDIV = 32 1 = 31 = 0x001f February 11, 2012 Veton Kpuska 82 SPI_CTL Register The SPI Control register (SPI_CTL) is used to configure and enable the SPI system. This register is used to enable the SPI interface, select the device as determine the data transfer a master or slave, and format and word size. February 11, 2012 Veton Kpuska 83 SPI_CTL Register Details The term "word" refers to a single data transfer of either 8 bits or 16 bits, depending on the word length (SIZE) bit in SPI_CTL. There are two special bits which can also be modified by the hardware: SPE and MSTR. The TIMOD field is used to specify the action that initiates transfers to/from the receive/transmit buffers. When set to 00, a SPI port transaction is begun when the receive buffer is read. Data from the first read will need to be discarded since the read is needed to initiate the first SPI port transaction. When set to 01, the transaction is initiated when the transmit buffer is written. A value of 10 selects DMA Receive Mode and the first transaction is initiated by enabling the SPI for DMA Receive mode. Subsequent individual transactions are initiated by a DMA read of the SPI_RDBR. A value of 11 selects DMA Transmit Mode and the transaction is initiated by a DMA write of the SPI_TDBR. February 11, 2012 Veton Kpuska 84 SPI_CTL Register Details The PSSE bit is used to enable the SPISS input for master. When not used, SPISS can be disabled, freeing up a chip pin as generalpurpose I/O. The EMISO bit enables the MISO pin as an output. This is needed in an environment where the master wishes to transmit to various slaves at one time (broadcast). Only one slave is allowed to transmit data back to the master. Except for the slave from whom the master wishes to receive, all other slaves should have this bit cleared. The SPE and MSTR bits can be modified by hardware when the MODF bit of the Status register is set. See "Mode Fault Error" on page 1027 of ADSPBF533 Blackfin Processor Hardware Reference manual. February 11, 2012 Veton Kpuska 85 SPI_CTL Register Details *pSPI_CTL == TIMOD_DMA_TX || SIZE || MSTR *pSPI_CTL TIMOD_DMA_TX SIZE MSTR TIMOD_DMA_TX 0x0003 == 0000 0000 0000 0011 TIMOD_DMA_TX 0x0003 0000 0000 0000 0011 February 11, 2012 Veton Kpuska 86 Summary of SPI Register Functions February 11, 2012 Veton Kpuska 87 BF533 Direct Memory Access DMA BF533 DMA Support Processor Memory Architecture February 11, 2012 Veton Kpuska 89 BF533 DMA Support The processor has multiple, independent DMA controllers that support automated data transfers with minimal overhead for the core. DMA transfers can occur between the internal memories and any of its DMAcapable peripherals. Additionally, DMA transfers can be accomplished between any of the DMAcapable peripherals and external devices connected to the external memory interfaces, including the SDRAM controller and the asynchronous memory controller. DMAcapable peripherals include the SPORTs, SPI port, UART, and PPI. Each individual DMAcapable peripheral has at least one dedicated DMA channel. The DMA controller supports both onedimensional (1D) and twodimensional (2D) DMA transfers. DMA transfer initialization can be implemented from registers or from sets of parameters called descriptor blocks. The 2D DMA capability supports arbitrary row and column sizes up to 64K elements by 64K elements, and arbitrary row and column step sizes up to +/ 32K elements. Furthermore, the column step size can be less than the row step size, allowing implementation of interleaved datastreams. This feature is especially useful in video applications where data can be deinterleaved on the fly. February 11, 2012 Veton Kpuska 90 Generic Names of the DMA Memory Mapped Registers February 11, 2012 Veton Kpuska 91 DMA Configuration 1. 2. 3. 4. 5. 6. 7. 8. 9. EBIU setup Flash setup FIO setup ADI1836 DMA SPORT0 DMA configuration SPORT IRS DMA Support Endless loop and processing Flag Interrupts Initialize.c: Init1836() // Set up DMA5 to transmit // Map DMA5 to SPI *pDMA5_PERIPHERAL_MAP = 0x5000; // Configure DMA5 // 16-bit transfers *pDMA5_CONFIG = WDSIZE_16; // Start address of data buffer *pDMA5_START_ADDR = sCodec1836TxRegs; // DMA inner loop count *pDMA5_X_COUNT = CODEC_1836_REGS_LENGTH; // Inner loop address increment *pDMA5_X_MODIFY = 2; // enable DMAs *pDMA5_CONFIG = (*pDMA5_CONFIG | DMAEN); // enable spi *pSPI_CTL = (*pSPI_CTL | SPE); // wait until dma transfers for spi are finished for (j=0; j<0xaff; j++); // disable spi *pSPI_CTL = 0x0000; } February 11, 2012 Veton Kpuska 93 DMA Channel's Peripheral Map Register Each DMA channel's Peripheral Map register (DMAx_PERIPHERAL_MAP) contains bits that: Map the channel to a specific peripheral. Identify whether the channel is a Peripheral DMA channel or a Memory DMA channel. *pDMA5_PERIPHERAL_MAP *pDMA5_PERIPHERAL_MAP = 0x5000; = 0x5000; February 11, 2012 Veton Kpuska 94 Initialize.c: Init1836() // Set up DMA5 to transmit // Map DMA5 to SPI *pDMA5_PERIPHERAL_MAP = 0x5000; // Configure DMA5 // 16-bit transfers *pDMA5_CONFIG = WDSIZE_16; // Start address of data buffer *pDMA5_START_ADDR = sCodec1836TxRegs; // DMA inner loop count *pDMA5_X_COUNT = CODEC_1836_REGS_LENGTH; // Inner loop address increment *pDMA5_X_MODIFY = 2; // enable DMAs *pDMA5_CONFIG = (*pDMA5_CONFIG | DMAEN); // enable spi *pSPI_CTL = (*pSPI_CTL | SPE); // wait until dma transfers for spi are finished for (j=0; j<0xaff; j++); // disable spi *pSPI_CTL = 0x0000; } February 11, 2012 Veton Kpuska 95 DMA Configuration Register The DMA Configuration register (DMAx_CONFIG/MDMA_yy_CONFIG), shown in Figure 93 of ADSPBF533 Blackfin Processor Hardware Reference, also in this slide is used to set up DMA parameters and operating modes. February 11, 2012 Veton Kpuska 96 Initialize.c: Init1836() // Set up DMA5 to transmit // Map DMA5 to SPI *pDMA5_PERIPHERAL_MAP = 0x5000; // Configure DMA5 // 16-bit transfers *pDMA5_CONFIG = WDSIZE_16; // Start address of data buffer *pDMA5_START_ADDR = sCodec1836TxRegs; // DMA inner loop count *pDMA5_X_COUNT = CODEC_1836_REGS_LENGTH; // Inner loop address increment *pDMA5_X_MODIFY = 2; // enable DMAs *pDMA5_CONFIG = (*pDMA5_CONFIG | DMAEN); // enable spi *pSPI_CTL = (*pSPI_CTL | SPE); // wait until dma transfers for spi are finished for (j=0; j<0xaff; j++); // disable spi *pSPI_CTL = 0x0000; } February 11, 2012 Veton Kpuska 97 DMA Start Address Register The Start Address register (DMAx_START_ADDR/MDMA_yy_START_ADDR), shown in Figure 92 of ADSPBF533 Blackfin Processor Hardware Reference, also in the next slide, contains the start address of the data buffer currently targeted for DMA. February 11, 2012 Veton Kpuska 98 Initialize.c: Init1836() // Set up DMA5 to transmit // Map DMA5 to SPI *pDMA5_PERIPHERAL_MAP = 0x5000; // Configure DMA5 // 16-bit transfers *pDMA5_CONFIG = WDSIZE_16; // Start address of data buffer *pDMA5_START_ADDR = sCodec1836TxRegs; // DMA inner loop count *pDMA5_X_COUNT = CODEC_1836_REGS_LENGTH; // Inner loop address increment *pDMA5_X_MODIFY = 2; // enable DMAs *pDMA5_CONFIG = (*pDMA5_CONFIG | DMAEN); // enable spi *pSPI_CTL = (*pSPI_CTL | SPE); // wait until dma transfers for spi are finished for (j=0; j<0xaff; j++); // disable spi *pSPI_CTL = 0x0000; } February 11, 2012 Veton Kpuska 99 DMA Count Register For 1D DMA, it specifies the number of elements to read in. For 2D DMA details, see "Two Dimensional DMA" on page 945 of ADSPBF533 Blackfin Processor Hardware Reference. A value of 0 in X_COUNT corresponds to 65,536 elements. February 11, 2012 Veton Kpuska 100 Initialize.c: Init1836() // Set up DMA5 to transmit // Map DMA5 to SPI *pDMA5_PERIPHERAL_MAP = 0x5000; // Configure DMA5 // 16-bit transfers *pDMA5_CONFIG = WDSIZE_16; // Start address of data buffer *pDMA5_START_ADDR = sCodec1836TxRegs; // DMA inner loop count *pDMA5_X_COUNT = CODEC_1836_REGS_LENGTH; // Inner loop address increment *pDMA5_X_MODIFY = 2; // enable DMAs *pDMA5_CONFIG = (*pDMA5_CONFIG | DMAEN); // enable spi *pSPI_CTL = (*pSPI_CTL | SPE); // wait until dma transfers for spi are finished for (j=0; j<0xaff; j++); // disable spi *pSPI_CTL = 0x0000; } February 11, 2012 Veton Kpuska 101 DMAx Modify Register The Inner Loop Address Increment register (DMAx_X_MODIFY/MDMA_yy_X_MODIFY) contains a signed, two'scomplement byteaddress increment. In 1D DMA, this increment is the stride that is applied after transferring each element. Note X_MODIFY is specified in bytes, regardless of the DMA transfer size. February 11, 2012 Veton Kpuska 102 Initialize.c: Init1836() // Set up DMA5 to transmit // Map DMA5 to SPI *pDMA5_PERIPHERAL_MAP = 0x5000; // Configure DMA5 // 16-bit transfers *pDMA5_CONFIG = WDSIZE_16; // Start address of data buffer *pDMA5_START_ADDR = sCodec1836TxRegs; // DMA inner loop count *pDMA5_X_COUNT = CODEC_1836_REGS_LENGTH; // Inner loop address increment *pDMA5_X_MODIFY = 2; // enable DMAs *pDMA5_CONFIG = (*pDMA5_CONFIG | DMAEN); // enable spi *pSPI_CTL = (*pSPI_CTL | SPE); // wait until dma transfers for spi are finished for (j=0; j<0xaff; j++); // disable spi *pSPI_CTL = 0x0000; } February 11, 2012 Veton Kpuska 103 DMA Configuration Register *pDMA5_CONFIG = WDSIZE_16 = 0x0004; *pDMA5_CONFIG = WDSIZE_16 = 0x0004; *pDMA5_CONFIG = (*pDMA5_CONFIG | DMAEN); *pDMA5_CONFIG = (*pDMA5_CONFIG | DMAEN); DMAEN = 0x0001 DMAEN = 0x0001 February 11, 2012 Veton Kpuska 104 SPI_CTL Register Details *pSPI_CTL == TIMOD_DMA_TX || SIZE || MSTR *pSPI_CTL TIMOD_DMA_TX SIZE MSTR TIMOD_DMA_TX 0x0003 == 0000 0000 0000 0011 TIMOD_DMA_TX 0x0003 0000 0000 0000 0011 *pSPI_CTL == *pSPI_CTL || SPE *pSPI_CTL *pSPI_CTL SPE SPE 0x4000 == 0100 0000 0000 0000 SPE 0x4000 0100 0000 0000 0000 February 11, 2012 Veton Kpuska 105 Serial Port Controllers: SPORT 1. 2. 3. 4. 5. 6. 7. 8. 9. EBIU setup Flash setup FIO setup ADI1836 DMA SPORT0 DMA configuration SPORT IRS DMA Support Endless loop and processing Flag Interrupts Serial Port Controllers (SPORT) The processor has two identical synchronous serial ports, or SPORTs. These support a variety of serial data communications protocols and can provide a direct interconnection between processors in a multiprocessor system. The serial ports (SPORT0 and SPORT1) provide an I/O interface to a wide variety of peripheral serial devices. SPORTs provide synchronous serial data transfer only; The processor provides asynchronous RS232 data transfer via the UART. Each SPORT has one group of pins: The receive and transmit functions are programmed separately. Each SPORT is a full duplex device, capable of simultaneous data transfer in both directions. The SPORTs can be programmed for primary data, secondary data, clock, and frame sync for transmit and a second set of pins for receive. bit rate, frame sync, and number of bits per word by writing to memorymapped registers. February 11, 2012 Veton Kpuska 107 SPORT (cont.) Both SPORTs have the same capabilities and are programmed in the same way. Each SPORT has its own set of control registers and data buffers. The SPORTs use frame sync pulses to indicate the beginning of each word or packet, and the bit clock marks the beginning of each data bit. External bit clock and frame sync are available for the TX and RX buffers. Veton Kpuska 108 February 11, 2012 SPORT (cont.) With a range of clock and frame synchronization options, the SPORTs allow a variety of serial communication protocols. The SPORTs can operate at up to an SCLK/2 clock rate with an externally generated clock, or 1/2 the system clock rate for an internally generated serial port clock. The SPORT external clock must always be less than the SCLK frequency. Independent transmit and receive clocks provide greater flexibility for serial communications. SPORT clocks and frame syncs can be internally generated by the system or received from an external source. The SPORTs can operate with a transmission format of LSB first or MSB first, with word lengths selectable from 3 to 32 bits. They offer selectable transmit modes and optional law or Alaw companding in hardware. SPORT data can be automatically transferred between onchip and offchip memories using DMA block transfers. Additionally, each of the SPORTs offers a TDM (TimeDivisionMultiplexed) Multichannel mode. February 11, 2012 Veton Kpuska 109 SPORT Features Provides independent transmit and receive functions. Transfers serial data words from 3 to 32 bits in length, either MSB first or LSB first. Provides alternate framing and control for interfacing to I2S serial devices, as well as other audio formats (for example, leftjustified stereo serial data). Has FIFO plus double buffered data (both receive and transmit functions have a data buffer register and a Shift register), providing additional time to service the SPORT. Provides two synchronous transmit and two synchronous receive data pins and buffers in each SPORT to double the total supported datastreams. February 11, 2012 Veton Kpuska 110 SPORT Features Performs Alaw and law hardware companding on transmitted and received words. (See "Companding" on page 1235 for more information.) Internally generates serial clock and frame sync signals in a wide range of frequencies or accepts clock and frame sync input from an external source. Operates with or without frame synchronization signals for each data word, with internally generated or externally generated frame signals, with active high or active low frame signals, and with either of two configurable pulse widths and frame signal timing. Performs interruptdriven, single word transfers to and from onchip memory under processor control. February 11, 2012 Veton Kpuska 111 SPORT Features Provides Direct Memory Access transfer to and from memory under DMA Master control. DMA can be autobufferbased (a repeated, identical range of transfers) or descriptorbased (individual or repeated ranges of transfers with differing DMA parameters). Executes DMA transfers to and from onchip memory. Each SPORT can automatically receive and transmit an entire block of data. Permits chaining of DMA operations for multiple data blocks. Has a multichannel mode for TDM interfaces. Each SPORT can receive and transmit data selectively from a TimeDivisionMultiplexed serial bitstream on 128 contiguous channels from a stream of up to 1024 total channels. This mode can be useful as a network communication scheme for multiple processors. The 128 channels available to the processor can be selected to start at any channel location from 0 to 895 = (1023 128). Note the Multichannel Select registers and the WSIZE register control which subset of the 128 channels within the active region can be accessed. February 11, 2012 Veton Kpuska 112 SPORT Pins February 11, 2012 Veton Kpuska 113 SPORT Data to be transmitted is written from an The receive portion of the SPORT accepts A SPORT For transmit: Data to be transmitted is written from an The receive portion of the SPORT accepts A SPORT For transmit: internal processor register to the SPORT's data from the data from the Receives serial data on its DRxPRI and internal processor register to the SPORT's The data bits (DTxPRI and DTxSEC) are Receives serial data on its DRxPRI and The data bits (DTxPRI and DTxSEC) are SPORTx_TX register via the peripheral bus. SPORTx_TX register via the peripheral bus. DRxSEC inputs and DRxPRI/DRxSEC pin synchronous to the transmit clock DRxSEC inputs and DRxPRI/DRxSEC pin synchronous to the transmit clock This data is optionally compressed by the (TSCLKx). This data is optionally compressed by the (TSCLKx). Transmits serial data on its DTxPRI and synchronous to the serial clock on the Transmits serial data on its DTxPRI and synchronous to the serial clock on the hardware and automatically transferred to hardware and automatically transferred to RSCLKx pin. DTxSEC outputs. For receive: RSCLKx pin. DTxSEC outputs. For receive: the TX Shift register. the TX Shift register. It can receive and transmit simultaneously When an entire word is received, the data The data bits (DRxPRI and DRxSEC) are It can receive and transmit simultaneously When an entire word is received, the data The data bits (DRxPRI and DRxSEC) are The bits in the Shift register are shifted out The bits in the Shift register are shifted out is optionally expanded, then for fullduplex operation. synchronous to the receive clock is optionally expanded, then for fullduplex operation. synchronous to the receive clock on the SPORT's on the SPORT's (RSCLKx). (RSCLKx). automatically transferred to the SPORT's automatically transferred to the SPORT's DTxPRI/DTxSEC pin, DTxPRI/DTxSEC pin, SPORTx_RX register, and then The serial clock is an output if the SPORTx_RX register, and then The serial clock is an output if the MSB first or LSB first, processor generates it, or an input if the MSB first or LSB first, processor generates it, or an input if the into the RX FIFO where it is available to into the RX FIFO where it is available to clock is externally generated. clock is externally generated. the processor. Synchronous to the serial clock on the processor. Synchronous to the serial clock on the TSCLKx pin. Frame synchronization signals RFSx and the TSCLKx pin. Frame synchronization signals RFSx and TFSx are used to indicate the start of a TFSx are used to indicate the start of a serial data word or stream of serial words. serial data word or stream of serial words. February 11, 2012 Veton Kpuska 114 SPORT Generic Operation Writing to a SPORT's SPORTx_TX register readies the SPORT for transmission. The TFS signal initiates the transmission of serial data. Once transmission has begun, each value written to the SPORTx_TX register is transferred through the FIFO to the internal Transmit Shift register. The bits are then sent, beginning with either the MSB or the LSB as specified in the SPORTx_TCR1 register. Each bit is shifted out on the driving edge of TSCLKx. The driving edge of TSCLKx can be configured to be rising or falling. the transmit interrupt or requests a DMA transfer as long as there is space in the TX FIFO. The SPORT generates As a SPORT receives bits, they accumulate in an internal receive register. When a complete word has been received, it is written to the SPORT FIFO register and the receive interrupt for that SPORT is generated or A DMA transfer is initiated. Interrupts are generated differently if DMA block transfers are performed. For information about DMA, see Chapter 9, "Direct Memory Access." February 11, 2012 Veton Kpuska 115 Generic Multichannel SPORT Connectivity February 11, 2012 Veton Kpuska 116 AD1836 and SPORT February 11, 2012 Veton Kpuska 117 Initialize.c : Init_Sport0 //--------------------------------------------------------------------------// // Function:Init_Sport0 // // // // Description: Configure Sport0 for TDM mode, to transmit/receive data // // to/from the AD1836. Configure Sport for external clocks and // // frame syncs. // //--------------------------------------------------------------------------// void Init_Sport0(void) { // Sport0 receive configuration // External CLK, External Frame sync, MSB first // 32-bit data *pSPORT0_RCR1 = RFSR; *pSPORT0_RCR2 = SLEN_32; // Sport0 transmit configuration // External CLK, External Frame sync, MSB first // 24-bit data *pSPORT0_TCR1 = TFSR; *pSPORT0_TCR2 = SLEN_32; // Enable MCM 8 transmit & receive channels *pSPORT0_MTCS0 = 0x000000FF; *pSPORT0_MRCS0 = 0x000000FF; // Set MCM configuration register and enable MCM mode *pSPORT0_MCMC1 = 0x0000; *pSPORT0_MCMC2 = 0x101c; } February 11, 2012 Veton Kpuska 118 Setting SPORT Modes SPORT configuration is accomplished by setting bit and field values in configuration registers. Each SPORT must be configured prior to being enabled. Once the SPORT is enabled, further writes to the SPORT Configuration registers are disabled (except for SPORTx_RCLKDIV, SPORTx_TCLKDIV, and Multichannel Mode Channel Select registers). To change values in all other SPORT Configuration registers, disable the SPORT by clearing TSPEN in SPORTx_TCR1 and/or RSPEN in SPORTx_RCR1. Each SPORT has its own set of control registers and data buffers. These registers are described in detail in the following sections as they relate to audio LAB's. All control and status bits in the SPORT registers are active high unless otherwise noted. Veton Kpuska 119 February 11, 2012 SPORTx_RCR1 and SPORTx_RCR2 Registers The main control registers for the receive portion of each SPORT are the Receive Configuration registers: SPORTx_RCR1 and SPORTx_RCR2. A SPORT is enabled for receive if Bit 0 (RSPEN) of the Receive Configuration 1 register is set to 1. This bit is cleared during either a hard reset or a soft reset, disabling all SPORT reception. When the SPORT is enabled to receive (RSPEN bit is set), corresponding SPORT Configuration register writes are not allowed except for SPORTx_RCLKDIV and Multichannel Mode Channel Select registers. Writes to disallowed registers have no effect. While the SPORT is enabled, SPORTx_RCR1 is not written except for bit 0 (RSPEN). For example: write (SPORTx_RCR1, 0x0001) ; /* SPORT RX Enabled */ write (SPORTx_RCR1, 0xFF01) ; /* ignored, no effect */ write (SPORTx_RCR1, 0xFFF0) ; /* SPORT disabled, SPORTx_RCR1 still equal to 0x0000 */ February 11, 2012 Veton Kpuska 120 SPORTx Receive Configuration 1 Register *pSPORT0_RCR1 = RFSR; *pSPORT0_RCR1 = RFSR; RFSR = 0x0000 or 0x0400; RFSR = 0x0000 or 0x0400; February 11, 2012 Veton Kpuska 121 SPORTx Receive Configuration 2 Register *pSPORT0_RCR2 = SLEN_32 *pSPORT0_RCR2 = SLEN_32 SLEN_32 = 0x001f; SLEN_32 = 0x001f; February 11, 2012 Veton Kpuska 122 SPORTx_RCR1 and SPORTx_RCR2 Registers Additional information for the SPORTx_RCR1 and SPORTxRCR2 Receive Configuration register bits: Receive Enable (RSPEN). This bit selects whether the SPORT is enabled to receive (if set) or disabled (if cleared). Setting the RSPEN bit turns on the SPORT and causes it to sample data from the data receive pins as well as the receive bit clock and Receive Frame Sync pins if so programmed. Setting RSPEN enables the SPORTx receiver, which can generate a SPORTx RX interrupt. For this reason, the code should initialize the ISR and the DMA control registers first, and should be ready to service RX interrupts before setting RSPEN. Setting RSPEN also generates DMA requests if DMA is enabled and data is received. Set all DMA control registers before setting RSPEN. February 11, 2012 Veton Kpuska 123 SPORTx_RCR1 and SPORTx_RCR2 Registers Clearing RSPEN causes the SPORT to stop receiving data; it also shuts down the internal SPORT receive circuitry. In low power applications, battery life can be extended by clearing RSPEN whenever the SPORT is not in use. Internal Receive Clock Select. (IRCLK). This bit selects the internal receive clock (if set) or external receive clock (if cleared). The RCLKDIV MMR value is not used when an external clock is selected. Data Formatting Type Select. (RDTYPE). The two RDTYPE bits specify one of four data formats used for single and multichannel operation. Bit Order Select. (RLSBIT). The RLSBIT bit selects the bit order of the data words received over the SPORTs. February 11, 2012 Veton Kpuska 124 SPORTx_RCR1 and SPORTx_RCR2 Registers Serial Word Length Select. (SLEN). The serial word length (the number of bits in each word received over the SPORTs) is calculated by adding 1 to the value of the SLEN field. The SLEN field can be set to a value of 2 to 31; 0 and 1 are illegal values for this field. Internal Receive Frame Sync Select. (IRFS). This bit selects whether the SPORT uses an internal RFS (if set) or an external RFS (if cleared). Receive Frame Sync Required Select. (RFSR). This bit selects whether the SPORT requires (if set) or does not require (if cleared) a Receive Frame Sync for every data word. Low Receive Frame Sync Select. (LRFS). This bit selects an active low RFS (if set) or active high RFS (if cleared). Late Receive Frame Sync. (LARFS). This bit configures late frame syncs (if set) or early frame syncs (if cleared). February 11, 2012 Veton Kpuska 125 SPORTx_RCR1 and SPORTx_RCR2 Registers Clock Drive/Sample Edge Select. (RCKFE). This bit selects which edge of the RSCLK clock signal the SPORT uses for sampling data, for sampling externally generated frame syncs, and for driving internally generated frame syncs. If set, internally generated frame syncs are driven on the falling edge, and data and externally generated frame syncs are sampled on the rising edge. If cleared, internally generated frame syncs are driven on the rising edge, and data and externally generated frame syncs are sampled on the falling edge. RxSec Enable. (RXSE). This bit enables the receive secondary side of the serial port (if set). Stereo Serial Enable. (RSFSE). This bit enables the Stereo Serial operating mode of the serial port (if set). By default this bit is cleared, enabling normal clocking and frame sync. Left/Right Order. (RRFST). If this bit is set, the right channel is received first in Stereo Serial operating mode. By default this bit is cleared, and the left channel is received first. February 11, 2012 Veton Kpuska 126 SPORTx_TCR1 and SPORTx_TCR2 Registers The main control registers for the transmit portion of each SPORT are the Transmit Configuration registers, SPORTx_TCR1 and SPORTx_TCR2. A SPORT is enabled for transmit if Bit 0 (TSPEN) of the Transmit Configuration 1 register is set to 1. This bit is cleared during either a hard reset or a soft reset, disabling all SPORT transmission. When the SPORT is enabled to transmit (TSPEN set), corresponding SPORT Configuration register writes are not allowed except for SPORTx_TCLKDIV and Multichannel Mode Channel Select registers. Writes to disallowed registers have no effect. While the SPORT is enabled, SPORTx_TCR1 is not written except for bit 0 (TSPEN). For example: write (SPORTx_TCR1, 0x0001) ; /* SPORT TX Enabled */ write (SPORTx_TCR1, 0xFF01) ; /* ignored, no effect */ write (SPORTx_TCR1, 0xFFF0) ; /* SPORT disabled, SPORTx_TCR1 still equal to 0x0000 */ February 11, 2012 Veton Kpuska 127 SPORTX Transmit Configuration 1 Register: SPORTx_TCR1 *pSPORT0_TCR1 = TFSR; *pSPORT0_TCR1 = TFSR; TFSR = 0x0000 or 0x0400; TFSR = 0x0000 or 0x0400; February 11, 2012 Veton Kpuska 128 SPORTX Transmit Configuration 2 Register: SPORTx_TCR2 *pSPORT0_TCR2 = SLEN_32; *pSPORT0_TCR2 = SLEN_32; SLEN_32 = 0x001f; SLEN_32 = 0x001f; February 11, 2012 Veton Kpuska 129 SPORTx_TCR1 and SPORTx_TCR2 Registers Additional information for the SPORTx_TCR1 and SPORTx_TCR2 Transmit Configuration register bits includes: Transmit Enable (TSPEN). This bit selects whether the SPORT is enabled to transmit (if set) or disabled (if cleared). Setting TSPEN causes an immediate assertion of a SPORT TX interrupt, indicating that the TX data register is empty and needs to be filled. This is normally desirable because it allows centralization of the transmit data write code in the TX interrupt service routine (ISR). For this reason, the code should initialize the ISR and be ready to service TX interrupts before setting TSPEN. Similarly, if DMA transfers will be used, DMA control should be configured correctly before setting TSPEN. Set all DMA control registers before setting TSPEN. Clearing TSPEN causes the SPORT to stop driving data, TSCLK, and frame sync pins; it also shuts down the internal SPORT circuitry. In low power applications, battery life can be extended by clearing TSPEN whenever the SPORT is not in use. All SPORT control registers should be programmed before TSPEN is set. Typical SPORT initialization code first writes all control registers, including DMA control if applicable. The last step in the code is to write SPORTx_TCR1 with all of the necessary bits, including TSPEN. February 11, 2012 Veton Kpuska 130 SPORTx_TCR1 and SPORTx_TCR2 Registers Internal Transmit Clock Select. (ITCLK). This bit selects the internal transmit clock (if set) or the external transmit clock on the TSCLK pin (if cleared). The TCLKDIV MMR value is not used when an external clock is selected. Data Formatting Type Select. The two TDTYPE bits specify data formats used for single and multichannel operation. Bit Order Select. (TLSBIT). The TLSBIT bit selects the bit order of the data words transmitted over the SPORT. Serial Word Length Select. (SLEN). The serial word length (the number of bits in each word transmitted over the SPORTs) is calculated by adding 1 to the value of the SLEN field: Serial Word Length = SLEN + 1; The SLEN field can be set to a value of 2 to 31; 0 and 1 are illegal values for this field. Three common settings for the SLEN field are 15, to transmit a full 16bit word; 7, to transmit an 8bit byte; and 23, to transmit a 24bit word. The processor can load 16 or 32bit values into the transmit buffer via DMA or an MMR write instruction; the SLEN field tells the SPORT how many of those bits to shift out of the register over the serial link. The serial port transfers bits [SLEN:0] from the transmit buffer. February 11, 2012 Veton Kpuska 131 SPORTx_TCR1 and SPORTx_TCR2 Registers Internal Transmit Frame Sync Select. (ITFS). This bit selects whether the SPORT uses an internal TFS (if set) or an external TFS (if cleared). Transmit Frame Sync Required Select. (TFSR). This bit selects whether the SPORT requires (if set) or does not require (if cleared) a Transmit Frame Sync for every data word. DataIndependent Transmit Frame Sync Select. (DITFS). This bit selects whether the SPORT generates a dataindependent TFS (sync at selected interval) or a data dependent TFS (sync when data is present in SPORTx_TX) for the case of internal frame sync select (ITFS = 1). The DITFS bit is ignored when external frame syncs are selected. The frame sync pulse marks the beginning of the data word. If DITFS is set, the frame sync pulse is issued on time, whether the SPORTx_TX register has been loaded or not; if DITFS is cleared, the frame sync pulse is only generated if the SPORTx_TX data register has been loaded. If the receiver demands regular frame sync pulses, DITFS should be set, and the processor should keep loading the SPORTx_TX register on time. If the receiver can tolerate occasional late frame sync pulses, DITFS should be cleared to prevent the SPORT from transmitting old data twice or transmitting garbled data if the processor is late in loading the SPORTx_TX register. February 11, 2012 Veton Kpuska 132 SPORTx_TCR1 and SPORTx_TCR2 Registers Low Transmit Frame Sync Select. (LTFS). This bit selects an active low TFS (if set) or active high TFS (if cleared). Late Transmit Frame Sync. (LATFS). This bit configures late frame syncs (if set) or early frame syncs (if cleared). Clock Drive/Sample Edge Select. (TCKFE). This bit selects which edge of the TCLKx signal the SPORT uses for driving data, for driving internally generated frame syncs, and for sampling externally generated frame syncs. If set, data and internally generated frame syncs are driven on the falling edge, and externally generated frame syncs are sample on the rising edge. If cleared, data and internally generated frame syncs are driven on the rising edge, and externally generated frame syncs are sampled on the falling edge. TxSec Enable. (TXSE). This bit enables the transmit secondary side of the serial port (if set). Stereo Serial Enable. (TSFSE). This bit enables the Stereo Serial operating mode of the serial port (if set). By default this bit is cleared, enabling normal clocking and frame sync. Left/Right Order. (TRFST). If this bit is set, the right channel is transmitted first in Stereo Serial operating mode. By default this bit is cleared, and the left channel is transmitted first. February 11, 2012 Veton Kpuska 133 Data Word Formats The format of the data words transferred over the SPORTs is configured by the combination of: transmit SLEN and receive SLEN; RDTYPE; TDTYPE; RLSBIT; and TLSBIT bits of the SPORTx_TCR1, SPORTx_TCR2, SPORTx_RCR1, and SPORTx_RCR2 registers. Veton Kpuska February 11, 2012 134 SPORTs Multichannel Operation In Multichannel mode, RSCLK can either be provided externally or generated internally by the SPORT, and it is used for both transmit and receive functions. Leave TSCLK disconnected if the SPORT is used only in multichannel mode. If RSCLK is externally or internally provided, it will be internally distributed to both the receiver and transmitter circuitry. Veton Kpuska 135 February 11, 2012 SPORTs Multichannel Operation Important Note: The SPORT Multichannel Transmit Select register and the SPORT Multichannel Receive Select register must be programmed before enabling SPORTx_TX or SPORTx_RX operation for Multichannel Mode. This is especially important in "DMA data unpacked mode," since SPORT FIFO operation begins immediately after RSPEN and TSPEN are set, enabling both RX and TX. The MCMEN bit (in SPORTx_MCMC2) must be enabled prior to enabling SPORTx_TX or SPORTx_RX operation. When disabling the SPORT from multichannel operation, first disable TXEN and then disable RXEN. Note both TXEN and RXEN must be disabled before reenabling. Disabling only TX or RX is not allowed. February 11, 2012 Veton Kpuska 136 SPORTs Multichannel Operation SPORTx_MTCSn Registers The Multichannel Selection registers are used to enable and disable individual channels. The four SPORTx Multichannel Transmit Select registers (SPORTx_MTCSn) specify the active transmit channels. There are four registers, each with 32 bits, corresponding to the 128 channels: Setting a bit enables that channel so that the serial port selects that word for transmit from the multiple word block of data. For example, setting bit 0 selects word 0, setting bit 12 selects word 12, and so on. February 11, 2012 Veton Kpuska 137 SPORTs Multichannel Operation Setting a particular bit in a SPORTx_MTCSn register causes the serial port to transmit the word in that channel's position of the datastream. Clearing the bit in the SPORTx_MTCSn register causes the serial port's data transmit pin to threestate during the time slot of that channel. February 11, 2012 Veton Kpuska 138 SPORTs Multichannel Operation SPORTx_MRCSn Registers The Multichannel Selection registers are used to enable and disable individual channels. The SPORTx Multichannel Receive Select registers (SPORTx_MRCSn) specify the active receive channels. There are four registers, each with 32 bits, corresponding to the 128 channels. Setting a bit enables that channel so that the serial port selects that word for receive from the multiple word block of data. For example, setting bit 0 selects word 0, setting bit 12 selects word 12, and so on. February 11, 2012 Veton Kpuska 139 SPORTx Multichannel Transmit Select Registers *pSPORT0_MTCS0 = 0x000000FF; *pSPORT0_MTCS0 = 0x000000FF; February 11, 2012 Veton Kpuska 140 SPORTx Multichannel Receive Select Registers *pSPORT0_MRCS0 = 0x000000FF; *pSPORT0_MRCS0 = 0x000000FF; February 11, 2012 Veton Kpuska 141 SPORTx_MCMCn Registers SPORTx_MCMCn Registers There are two SPORTx Multichannel Configuration registers (SPORTx_MCMCn) for each SPORT. The SPORTx_MCMCn registers are used to configure the multichannel operation of the SPORT. The two control registers are shown below. February 11, 2012 Veton Kpuska 142 SPORTx_MCMCn Registers *pSPORT0_MCMC1 = 0x0000; *pSPORT0_MCMC1 = 0x0000; Desired Window Size = 8 Desired Window Size = 8 February 11, 2012 Veton Kpuska 143 SPORTx_MCMCn Registers *pSPORT0_MCMC2 = 0x101c; *pSPORT0_MCMC2 = 0x101c; 0x101c = 0001 0000 0001 1100 0x101c = 0001 0000 0001 1100 February 11, 2012 Veton Kpuska 144 Initialize.c 1. 2. 3. 4. 5. 6. 7. 8. 9. EBIU setup Flash setup FIO setup ADI1836 DMA SPORT0 DMA Configuration SPORT IRS DMA Support Endless loop and processing Flag Interrupts DMA Configuration Init_DMA() Init_DMA() //--------------------------------------------------------------------------// // Function: Init_DMA // // // // Description: Initialize DMA1 in autobuffer mode to receive and DMA2 in // // autobuffer mode to transmit // //--------------------------------------------------------------------------// void Init_DMA(void) { Setting up DMA1SPORT0 as // Set up DMA1 to receive Setting up DMA1SPORT0 as // Map DMA1 to Sport0 RX Input/Receive Channel Input/Receive Channel *pDMA1_PERIPHERAL_MAP = 0x1000; // Configure DMA1 // 32-bit transfers, Interrupt on completion, Autobuffer mode *pDMA1_CONFIG = WNR | WDSIZE_32 | DI_EN | FLOW_1; // Start address of data buffer *pDMA1_START_ADDR = iRxBuffer1; // DMA inner loop count *pDMA1_X_COUNT = 8; // Inner loop address increment *pDMA1_X_MODIFY = 4; // Set up DMA2 to transmit // Map DMA2 to Sport0 TX *pDMA2_PERIPHERAL_MAP = 0x2000; // Configure DMA2 // 32-bit transfers, Autobuffer mode *pDMA2_CONFIG = WDSIZE_32 | FLOW_1; // Start address of data buffer *pDMA2_START_ADDR = iTxBuffer1; // DMA inner loop count *pDMA2_X_COUNT = 8; // Inner loop address increment *pDMA2_X_MODIFY = 4; } February 11, 2012 Veton Kpuska 147 DMA Channel's Peripheral Map Register Each DMA channel's Peripheral Map register (DMAx_PERIPHERAL_MAP) contains bits that: Map the channel to a specific peripheral. Identify whether the channel is a Peripheral DMA channel or a Memory DMA channel. *pDMA1_PERIPHERAL_MAP *pDMA1_PERIPHERAL_MAP = 0x1000; = 0x1000; February 11, 2012 Veton Kpuska 148 Initialize.c : Init_DMA() //--------------------------------------------------------------------------// // Function: Init_DMA // // // // Description: Initialize DMA1 in autobuffer mode to receive and DMA2 in // // autobuffer mode to transmit // //--------------------------------------------------------------------------// void Init_DMA(void) { // Set up DMA1 to receive // Map DMA1 to Sport0 RX *pDMA1_PERIPHERAL_MAP = 0x1000; // Configure DMA1 // 32-bit transfers, Interrupt on completion, Autobuffer mode *pDMA1_CONFIG = WNR | WDSIZE_32 | DI_EN | FLOW_1; // Start address of data buffer *pDMA1_START_ADDR = iRxBuffer1; // DMA inner loop count *pDMA1_X_COUNT = 8; // Inner loop address increment *pDMA1_X_MODIFY = 4; // Set up DMA2 to transmit // Map DMA2 to Sport0 TX *pDMA2_PERIPHERAL_MAP = 0x2000; // Configure DMA2 // 32-bit transfers, Autobuffer mode *pDMA2_CONFIG = WDSIZE_32 | FLOW_1; // Start address of data buffer *pDMA2_START_ADDR = iTxBuffer1; // DMA inner loop count *pDMA2_X_COUNT = 8; // Inner loop address increment *pDMA2_X_MODIFY = 4; } February 11, 2012 Veton Kpuska 149 DMA Configuration Register The DMA Configuration register (DMAx_CONFIG/MDMA_yy_CONFIG), shown in Figure 93 of ADSPBF533 Blackfin Processor Hardware Reference, also in this slide is used to set up DMA parameters and operating modes. *pDMA1_CONFIG = WNR | WDSIZE_32 | DI_EN | FLOW_1; *pDMA1_CONFIG = WNR | WDSIZE_32 | DI_EN | FLOW_1; February 11, 2012 Veton Kpuska 150 Initialize.c : Init_DMA() //--------------------------------------------------------------------------// // Function: Init_DMA // // // // Description: Initialize DMA1 in autobuffer mode to receive and DMA2 in // // autobuffer mode to transmit // //--------------------------------------------------------------------------// void Init_DMA(void) { // Set up DMA1 to receive // Map DMA1 to Sport0 RX *pDMA1_PERIPHERAL_MAP = 0x1000; // Configure DMA1 // 32-bit transfers, Interrupt on completion, Autobuffer mode *pDMA1_CONFIG = WNR | WDSIZE_32 | DI_EN | FLOW_1; // Start address of data buffer *pDMA1_START_ADDR = iRxBuffer1; // DMA inner loop count *pDMA1_X_COUNT = 8; // Inner loop address increment *pDMA1_X_MODIFY = 4; // Set up DMA2 to transmit // Map DMA2 to Sport0 TX *pDMA2_PERIPHERAL_MAP = 0x2000; // Configure DMA2 // 32-bit transfers, Autobuffer mode *pDMA2_CONFIG = WDSIZE_32 | FLOW_1; // Start address of data buffer *pDMA2_START_ADDR = iTxBuffer1; // DMA inner loop count *pDMA2_X_COUNT = 8; // Inner loop address increment *pDMA2_X_MODIFY = 4; } February 11, 2012 Veton Kpuska 151 DMA Start Address Register The Start Address register (DMAx_START_ADDR/MDMA_yy_START_ADDR), shown in Figure 92 of ADSPBF533 Blackfin Processor Hardware Reference, also in the next slide, contains the start address of the data buffer currently targeted for DMA. *pDMA1_START_ADDR = iRxBuffer1; *pDMA1_START_ADDR = iRxBuffer1; February 11, 2012 Veton Kpuska 152 Initialize.c : Init_DMA() //--------------------------------------------------------------------------// // Function: Init_DMA // // // // Description: Initialize DMA1 in autobuffer mode to receive and DMA2 in // // autobuffer mode to transmit // //--------------------------------------------------------------------------// void Init_DMA(void) { // Set up DMA1 to receive // Map DMA1 to Sport0 RX *pDMA1_PERIPHERAL_MAP = 0x1000; // Configure DMA1 // 32-bit transfers, Interrupt on completion, Autobuffer mode *pDMA1_CONFIG = WNR | WDSIZE_32 | DI_EN | FLOW_1; // Start address of data buffer *pDMA1_START_ADDR = iRxBuffer1; // DMA inner loop count *pDMA1_X_COUNT = 8; // Inner loop address increment *pDMA1_X_MODIFY = 4; // Set up DMA2 to transmit // Map DMA2 to Sport0 TX *pDMA2_PERIPHERAL_MAP = 0x2000; // Configure DMA2 // 32-bit transfers, Autobuffer mode *pDMA2_CONFIG = WDSIZE_32 | FLOW_1; // Start address of data buffer *pDMA2_START_ADDR = iTxBuffer1; // DMA inner loop count *pDMA2_X_COUNT = 8; // Inner loop address increment *pDMA2_X_MODIFY = 4; } February 11, 2012 Veton Kpuska 153 DMA Count Register For 1D DMA, it specifies the number of elements to read in. For 2D DMA details, see "Two Dimensional DMA" on page 945 of ADSPBF533 Blackfin Processor Hardware Reference. A value of 0 in X_COUNT corresponds to 65,536 elements. *pDMA1_X_COUNT = 8; *pDMA1_X_COUNT = 8; February 11, 2012 Veton Kpuska 154 Initialize.c : Init_DMA() //--------------------------------------------------------------------------// // Function: Init_DMA // // // // Description: Initialize DMA1 in autobuffer mode to receive and DMA2 in // // autobuffer mode to transmit // //--------------------------------------------------------------------------// void Init_DMA(void) { // Set up DMA1 to receive // Map DMA1 to Sport0 RX *pDMA1_PERIPHERAL_MAP = 0x1000; // Configure DMA1 // 32-bit transfers, Interrupt on completion, Autobuffer mode *pDMA1_CONFIG = WNR | WDSIZE_32 | DI_EN | FLOW_1; // Start address of data buffer *pDMA1_START_ADDR = iRxBuffer1; // DMA inner loop count *pDMA1_X_COUNT = 8; // Inner loop address increment *pDMA1_X_MODIFY = 4; // Set up DMA2 to transmit // Map DMA2 to Sport0 TX *pDMA2_PERIPHERAL_MAP = 0x2000; // Configure DMA2 // 32-bit transfers, Autobuffer mode *pDMA2_CONFIG = WDSIZE_32 | FLOW_1; // Start address of data buffer *pDMA2_START_ADDR = iTxBuffer1; // DMA inner loop count *pDMA2_X_COUNT = 8; // Inner loop address increment *pDMA2_X_MODIFY = 4; } February 11, 2012 Veton Kpuska 155 DMAx Modify Register The Inner Loop Address Increment register (DMAx_X_MODIFY/MDMA_yy_X_MODIFY) contains a signed, two'scomplement byteaddress increment. In 1D DMA, this increment is the stride that is applied after transferring each element. Note X_MODIFY is specified in bytes, regardless of the DMA transfer size. *pDMA1_X_MODIFY = 4; *pDMA1_X_MODIFY = 4; February 11, 2012 Veton Kpuska 156 Init_DMA() //--------------------------------------------------------------------------// // Function: Init_DMA // // // // Description: Initialize DMA1 in autobuffer mode to receive and DMA2 in // // autobuffer mode to transmit // //--------------------------------------------------------------------------// void Init_DMA(void) { // Set up DMA1 to receive // Map DMA1 to Sport0 RX *pDMA1_PERIPHERAL_MAP = 0x1000; // Configure DMA1 // 32-bit transfers, Interrupt on completion, Autobuffer mode *pDMA1_CONFIG = WNR | WDSIZE_32 | DI_EN | FLOW_1; // Start address of data buffer *pDMA1_START_ADDR = iRxBuffer1; // DMA inner loop count *pDMA1_X_COUNT = 8; // Inner loop address increment *pDMA1_X_MODIFY = 4; // Set up DMA2 to transmit // Map DMA2 to Sport0 TX *pDMA2_PERIPHERAL_MAP = 0x2000; // Configure DMA2 // 32-bit transfers, Autobuffer mode *pDMA2_CONFIG = WDSIZE_32 | FLOW_1; // Start address of data buffer *pDMA2_START_ADDR = iTxBuffer1; // DMA inner loop count *pDMA2_X_COUNT = 8; // Inner loop address increment *pDMA2_X_MODIFY = 4; } Setting up DMA2SPORT0 as Setting up DMA2SPORT0 as Output/Transmit Channel Output/Transmit Channel February 11, 2012 Veton Kpuska 157 DMA Channel's Peripheral Map Register Each DMA channel's Peripheral Map register (DMAx_PERIPHERAL_MAP) contains bits that: Map the channel to a specific peripheral. Identify whether the channel is a Peripheral DMA channel or a Memory DMA channel. *pDMA2_PERIPHERAL_MAP *pDMA2_PERIPHERAL_MAP = 0x2000; = 0x2000; February 11, 2012 Veton Kpuska 158 Initialize.c : Init_DMA() //--------------------------------------------------------------------------// // Function: Init_DMA // // // // Description: Initialize DMA1 in autobuffer mode to receive and DMA2 in // // autobuffer mode to transmit // //--------------------------------------------------------------------------// void Init_DMA(void) { // Set up DMA1 to receive // Map DMA1 to Sport0 RX *pDMA1_PERIPHERAL_MAP = 0x1000; // Configure DMA1 // 32-bit transfers, Interrupt on completion, Autobuffer mode *pDMA1_CONFIG = WNR | WDSIZE_32 | DI_EN | FLOW_1; // Start address of data buffer *pDMA1_START_ADDR = iRxBuffer1; // DMA inner loop count *pDMA1_X_COUNT = 8; // Inner loop address increment *pDMA1_X_MODIFY = 4; // Set up DMA2 to transmit // Map DMA2 to Sport0 TX *pDMA2_PERIPHERAL_MAP = 0x2000; // Configure DMA2 // 32-bit transfers, Autobuffer mode *pDMA2_CONFIG = WDSIZE_32 | FLOW_1; // Start address of data buffer *pDMA2_START_ADDR = iTxBuffer1; // DMA inner loop count *pDMA2_X_COUNT = 8; // Inner loop address increment *pDMA2_X_MODIFY = 4; } February 11, 2012 Veton Kpuska 159 DMA Configuration Register The DMA Configuration register (DMAx_CONFIG/MDMA_yy_CONFIG), shown in Figure 93 of ADSPBF533 Blackfin Processor Hardware Reference, also in this slide is used to set up DMA parameters and operating modes. *pDMA2_CONFIG = WDSIZE_32 | FLOW_1; *pDMA2_CONFIG = WDSIZE_32 | FLOW_1; February 11, 2012 Veton Kpuska 160 Initialize.c : Init_DMA() //--------------------------------------------------------------------------// // Function: Init_DMA // // // // Description: Initialize DMA1 in autobuffer mode to receive and DMA2 in // // autobuffer mode to transmit // //--------------------------------------------------------------------------// void Init_DMA(void) { // Set up DMA1 to receive // Map DMA1 to Sport0 RX *pDMA1_PERIPHERAL_MAP = 0x1000; // Configure DMA1 // 32-bit transfers, Interrupt on completion, Autobuffer mode *pDMA1_CONFIG = WNR | WDSIZE_32 | DI_EN | FLOW_1; // Start address of data buffer *pDMA1_START_ADDR = iRxBuffer1; // DMA inner loop count *pDMA1_X_COUNT = 8; // Inner loop address increment *pDMA1_X_MODIFY = 4; // Set up DMA2 to transmit // Map DMA2 to Sport0 TX *pDMA2_PERIPHERAL_MAP = 0x2000; // Configure DMA2 // 32-bit transfers, Autobuffer mode *pDMA2_CONFIG = WDSIZE_32 | FLOW_1; // Start address of data buffer *pDMA2_START_ADDR = iTxBuffer1; // DMA inner loop count *pDMA2_X_COUNT = 8; // Inner loop address increment *pDMA2_X_MODIFY = 4; } February 11, 2012 Veton Kpuska 161 DMA Start Address Register The Start Address register (DMAx_START_ADDR/MDMA_yy_START_ADDR), shown in Figure 92 of ADSPBF533 Blackfin Processor Hardware Reference, also in the next slide, contains the start address of the data buffer currently targeted for DMA. *pDMA2_START_ADDR = iTxBuffer1; *pDMA2_START_ADDR = iTxBuffer1; February 11, 2012 Veton Kpuska 162 Initialize.c : Init_DMA() //--------------------------------------------------------------------------// // Function: Init_DMA // // // // Description: Initialize DMA1 in autobuffer mode to receive and DMA2 in // // autobuffer mode to transmit // //--------------------------------------------------------------------------// void Init_DMA(void) { // Set up DMA1 to receive // Map DMA1 to Sport0 RX *pDMA1_PERIPHERAL_MAP = 0x1000; // Configure DMA1 // 32-bit transfers, Interrupt on completion, Autobuffer mode *pDMA1_CONFIG = WNR | WDSIZE_32 | DI_EN | FLOW_1; // Start address of data buffer *pDMA1_START_ADDR = iRxBuffer1; // DMA inner loop count *pDMA1_X_COUNT = 8; // Inner loop address increment *pDMA1_X_MODIFY = 4; // Set up DMA2 to transmit // Map DMA2 to Sport0 TX *pDMA2_PERIPHERAL_MAP = 0x2000; // Configure DMA2 // 32-bit transfers, Autobuffer mode *pDMA2_CONFIG = WDSIZE_32 | FLOW_1; // Start address of data buffer *pDMA2_START_ADDR = iTxBuffer1; // DMA inner loop count *pDMA2_X_COUNT = 8; // Inner loop address increment *pDMA2_X_MODIFY = 4; } February 11, 2012 Veton Kpuska 163 DMA Count Register For 1D DMA, it specifies the number of elements to read in. For 2D DMA details, see "Two Dimensional DMA" on page 945 of ADSPBF533 Blackfin Processor Hardware Reference. A value of 0 in X_COUNT corresponds to 65,536 elements. *pDMA2_X_COUNT = 8; *pDMA2_X_COUNT = 8; February 11, 2012 Veton Kpuska 164 Initialize.c : Init_DMA() //--------------------------------------------------------------------------// // Function: Init_DMA // // // // Description: Initialize DMA1 in autobuffer mode to receive and DMA2 in // // autobuffer mode to transmit // //--------------------------------------------------------------------------// void Init_DMA(void) { // Set up DMA1 to receive // Map DMA1 to Sport0 RX *pDMA1_PERIPHERAL_MAP = 0x1000; // Configure DMA1 // 32-bit transfers, Interrupt on completion, Autobuffer mode *pDMA1_CONFIG = WNR | WDSIZE_32 | DI_EN | FLOW_1; // Start address of data buffer *pDMA1_START_ADDR = iRxBuffer1; // DMA inner loop count *pDMA1_X_COUNT = 8; // Inner loop address increment *pDMA1_X_MODIFY = 4; // Set up DMA2 to transmit // Map DMA2 to Sport0 TX *pDMA2_PERIPHERAL_MAP = 0x2000; // Configure DMA2 // 32-bit transfers, Autobuffer mode *pDMA2_CONFIG = WDSIZE_32 | FLOW_1; // Start address of data buffer *pDMA2_START_ADDR = iTxBuffer1; // DMA inner loop count *pDMA2_X_COUNT = 8; // Inner loop address increment *pDMA2_X_MODIFY = 4; } February 11, 2012 Veton Kpuska 165 Initialize.c 1. 2. 3. 4. 5. 6. 7. 8. 9. EBIU setup Flash setup FIO setup ADI1836 DMA SPORT0 DMA Configuration SPORT IRS Enable DMA Support Endless loop and processing Flag Interrupts DMA Configuration Init_DMA() DMAx Modify Register The Inner Loop Address Increment register (DMAx_X_MODIFY/MDMA_yy_X_MODIFY) contains a signed, two'scomplement byteaddress increment. In 1D DMA, this increment is the stride that is applied after transferring each element. Note X_MODIFY is specified in bytes, regardless of the DMA transfer size. *pDMA2_X_MODIFY = 4; *pDMA2_X_MODIFY = 4; February 11, 2012 Veton Kpuska 168 Initialize.c : Enable_DMA_Sport() //--------------------------------------------------------------------------// // Function: Enable_DMA_Sport // // // // Description: Enable DMA1, DMA2, Sport0 TX and Sport0 RX // //--------------------------------------------------------------------------// void Enable_DMA_Sport0(void) { // enable DMAs *pDMA2_CONFIG = (*pDMA2_CONFIG | DMAEN); *pDMA1_CONFIG = (*pDMA1_CONFIG | DMAEN); // enable Sport0 TX and RX *pSPORT0_TCR1 = (*pSPORT0_TCR1 | TSPEN); *pSPORT0_RCR1 = (*pSPORT0_RCR1 | RSPEN); } February 11, 2012 Veton Kpuska 169 DMA Configuration Register *pDMA2_CONFIG = WDSIZE_32 | FLOW_1; *pDMA2_CONFIG = WDSIZE_32 | FLOW_1; *pDMA2_CONFIG = *pDMA2_CONFIG | DMAEN *pDMA2_CONFIG = *pDMA2_CONFIG | DMAEN *pDMA1_CONFIG = WNR | WDSIZE_32 | DI_EN | FLOW_1; *pDMA1_CONFIG = WNR | WDSIZE_32 | DI_EN | FLOW_1; *pDMA1_CONFIG = *pDMA1_CONFIG | DMAEN; *pDMA1_CONFIG = *pDMA1_CONFIG | DMAEN; February 11, 2012 Veton Kpuska 170 SPORTX Transmit Configuration 1 Register: SPORTx_TCR1 *pSPORT0_TCR1 = TFSR; *pSPORT0_TCR1 = TFSR; TFSR = 0x0000; TFSR = 0x0000; *pSPORT0_TCR1 *pSPORT0_TCR1 = (*pSPORT0_TCR1 | TSPEN); = (*pSPORT0_TCR1 | TSPEN); February 11, 2012 Veton Kpuska 171 SPORTx Receive Configuration 1 Register : SPORTx_RCR1 *pSPORT0_RCR1 = RFSR; *pSPORT0_RCR1 = RFSR; RFSR = 0x0000; RFSR = 0x0000; *pSPORT0_RCR1 *pSPORT0_RCR1 = (*pSPORT0_RCR1 | RSPEN); = (*pSPORT0_RCR1 | RSPEN); February 11, 2012 Veton Kpuska 172 Initialize.c 1. 2. 3. 4. 5. 6. 7. 8. 9. EBIU setup Flash setup FIO setup ADI1836 DMA SPORT0 DMA Configuration SPORT IRS Enable DMA Support Endless loop and processing Flag Interrupts Event Controller for Interrupts and Exceptions Event Controller The Event Controller of the processor manages five types of activities or events: Emulation Reset Nonmaskable interrupts (NMI) Exceptions Interrupts (11) Note the word event describes all five types of activities. The Event Controller manages fifteen different events in all: Emulation, Reset, NMI, Exception, and eleven Interrupts. Veton Kpuska 175 February 11, 2012 Event Controller An interrupt is an event that changes normal processor instruction flow and is asynchronous to program flow. In contrast, an exception is a software initiated event whose effects are synchronous to program flow. The event system is nested and prioritized. Consequently, several service routines may be active at any time, and a low priority event may be preempted by one of higher priority. February 11, 2012 Veton Kpuska 176 Event Controller The processor employs a twolevel event control mechanism. The SIC provides mapping between the many peripheral interrupt sources and the prioritized generalpurpose interrupt inputs of the core. The processor System Interrupt Controller (SIC) works with The Core Event Controller (CEC) to prioritize and control all system interrupts. This mapping is programmable, and individual interrupt sources can be masked in the SIC. The CEC supports nine (9) generalpurpose interrupts (IVG7 IVG15) in addition to the dedicated interrupt and exception events that are described in Table 46. It is recommended that the Refer to Table 46 for details. two lowest priority interrupts (IVG14 and IVG15) be reserved for software interrupt handlers, leaving seven prioritized interrupt inputs (IVG7 IVG13) to support the system. February 11, 2012 Veton Kpuska 177 Core Event Mapping February 11, 2012 Veton Kpuska 178 System Event Mapping February 11, 2012 Veton Kpuska 179 Processor Event Controller Processor Event Controller Consists of 2 stages: The Core Event Controller (CEC) System Interrupt Controller (SIC) Conceptually: Interrupts from the peripherals arrive at SIC SIC routes interrupts directly to general purpose interrupts of the CEC. February 11, 2012 Veton Kpuska 180 Core Event Controller (CEC) CEC supports 9 generalpurpose interrupts: IVG157 Additional dedicated interrupt and exception events. IVG1514 (2) lowest priority interrupts for software handlers. IRVG137 (7) highest to support peripherals. February 11, 2012 Veton Kpuska 181 System Interrupt Controller (SIC) SIC provides mapping and routing of events: Note: Processor default mapping can be altered by the user via Interrupt Assignment Register (IAR). From: Peripheral interrupt sources To: Prioritized generalpurpose interrupt inputs of the CEC. February 11, 2012 Veton Kpuska 182 BF533 System & Core Interrupt Controllers System Interrupt Source PLL Wakeup interrupt DMA error (generic) PPI error interrupt SPORT0 error interrupt SPORT1 error interrupt SPI error interrupt UART error interrupt RTC interrupt DMA 0 interrupt (PPI) DMA 1 interrupt (SPORT0 RX) DMA 2 interrupt (SPORT0 TX) DMA 3 interrupt (SPORT1 RX) DMA 4 interrupt (SPORT1 TX) DMA 5 interrupt (SPI) DMA 6 interrupt (UART RX) DMA 7 interrupt (UART TX) Timer0 interrupt Timer1 interrupt Timer2 interrupt PF interrupt A PF interrupt B DMA 8/9 interrupt (MemDMA0) DMA 10/11 interrupt (MemDMA1) Watchdog Timer Interrupt IVG # IVG7 IVG7 IVG7 IVG7 IVG7 IVG7 IVG7 IVG8 IVG8 IVG9 IVG9 IVG9 IVG9 IVG10 IVG10 IVG10 IVG11 IVG11 IVG11 IVG12 IVG12 IVG13 IVG13 IVG13 1 Event Source Emulator Reset Non Maskable Interrupt Exceptions Reserved Hardware Error Core Timer General Purpose 7 General Purpose 8 General Purpose 9 General Purpose 10 General Purpose 11 General Purpose 12 General Purpose 13 General Purpose 14 General Purpose 15 IVG # 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Core Event Name EMU RST NMI EVSW IVHW IVTMR IVG7 IVG8 IVG9 IVG10 IVG11 IVG12 IVG13 IVG14 IVG15 Highest Lowest February 11, 2012 Veton Kpuska 183 Priority System Interrupt Processing Referring to Figure 45, note when an interrupt (Interrupt A) is generated by an interruptenabled peripheral: SIC_ISR logs the request and keeps track of system interrupts that are asserted but not yet serviced (that is, an interrupt service routine hasn't yet cleared the interrupt). SIC_IWR checks to see if it should wake up the core from an idled state based on this interrupt request. SIC_IMASK masks off or enables interrupts from peripherals at the system level. If Interrupt A is not masked, the request proceeds to Step 4. The SIC_IARx registers, which map the peripheral interrupts to a smaller set of generalpurpose core interrupts (IVG7 IVG15), determine the core priority of Interrupt A. 1. 2. 3. 4. February 11, 2012 Veton Kpuska 184 System Interrupt Processing (cont.) 5. ILAT adds Interrupt A to its log of interrupts latched by the core but not yet actively being serviced. IMASK masks off or enables events of different core priorities. If the IVGx event corresponding to Interrupt A is not masked, the process proceeds to Step 7. The Event Vector Table (EVT) is accessed to look up the appropriate vector for Interrupt A's interrupt service routine (ISR). When the event vector for Interrupt A has entered the core pipeline, the appropriate IPEND bit is set, which clears the respective ILAT bit. Thus, IPEND tracks all pending interrupts, as well as those being presently serviced. When the interrupt service routine (ISR) for Interrupt A has been executed, the RTI instruction clears the appropriate IPEND bit. However, the relevant SIC_ISR bit is not cleared unless the interrupt service routine clears the mechanism that generated Interrupt A, or if the process of servicing the interrupt clears this bit. 6. 7. 8. 9. February 11, 2012 Veton Kpuska 185 Interrupt Processing Block Diagram It should be noted that emulation, reset, NMI, and exception events, as 8. When the event vector for Interrupt A has entered the core 9. When the interrupt service routine (ISR) for Interrupt A has been 7. The Event Vector Table (EVT) is accessed to look up the 5. ILAT adds Interrupt A to its log of interrupts latched by the core 6. IMASK masks off or enables events of different core priorities. If 4. The SIC_IARx registers, which map the peripheral interrupts to a 1. SIC_ISR logs the request and keeps track of system interrupts 3. SIC_IMASK masks off or enables interrupts from peripherals at 2. SIC_IWR checks to see if it should wake up the core from an It should be noted that emulation, reset, NMI, and exception events, as 8. The Event Vector Table (EVT) is accessed to look up the When the event vector for Interrupt A has entered the core 9. When the interrupt service routine (ISR) for Interrupt A has been 7. executed, the RTI instruction clears the appropriate IPEND bit. 5. pipeline, the appropriate IPEND bit is set, which clears the ILAT adds Interrupt A to its log of interrupts latched by the core 6. appropriate vector for Interrupt A's interrupt service routine (ISR). IMASK masks off or enables events of different core priorities. If 4. but not yet actively being serviced. The SIC_IARx registers, which map the peripheral interrupts to a 1. the IVGx event corresponding to Interrupt A is not masked, the SIC_ISR logs the request and keeps track of system interrupts 3. smaller set of generalpurpose core interrupts (IVG7 IVG15), SIC_IMASK masks off or enables interrupts from peripherals at 2. that are asserted but not yet serviced (that is, an interrupt SIC_IWR checks to see if it should wake up the core from an well as hardware error (IVHW) and core timer (IVTMR) interrupt the system level. If Interrupt A is not masked, the request idled state based on this interrupt request. pipeline, the appropriate IPEND bit is set, which clears the executed, the RTI instruction clears the appropriate IPEND bit. appropriate vector for Interrupt A's interrupt service routine (ISR). but not yet actively being serviced. the IVGx event corresponding to Interrupt A is not masked, the well as hardware error (IVHW) and core timer (IVTMR) interrupt smaller set of generalpurpose core interrupts (IVG7 IVG15), that are asserted but not yet serviced (that is, an interrupt the system level. If Interrupt A is not masked, the request idled state based on this interrupt request. respective ILAT bit. Thus, IPEND tracks all pending interrupts, as However, the relevant SIC_ISR bit is not cleared unless the process proceeds to Step 7. determine the core priority of Interrupt A. service routine hasn't yet cleared the interrupt). proceeds to Step 4. requests, enter the interrupt processing chain at the ILAT level and respective ILAT bit. Thus, IPEND tracks all pending interrupts, as However, the relevant SIC_ISR bit is not cleared unless the process proceeds to Step 7. determine the core priority of Interrupt A. service routine hasn't yet cleared the interrupt). proceeds to Step 4. requests, enter the interrupt processing chain at the ILAT level and well as those being presently serviced. interrupt service routine clears the mechanism that generated well as those being presently serviced. interrupt service routine clears the mechanism that generated are not affected by the systemlevel interrupt registers (SIC_IWR, Interrupt A, or if the process of servicing the interrupt clears this are not affected by the systemlevel interrupt registers (SIC_IWR, Interrupt A, or if the process of servicing the interrupt clears this SIC_ISR, SIC_IMASK, SIC_IARx). bit. SIC_ISR, SIC_IMASK, SIC_IARx). bit. February 11, 2012 Veton Kpuska 186 Interrupt Service Routine ISR address is stored in the Event Vector Table Used as the next fetch address when the event occurs Program Counter (PC) address is saved to a register RETI, RETX, RETN, RETE, based on event ISR always concludes with "Return" Instruction RTI, RTX, RTN, RTE (respectively) When executed, PC is loaded with address stored in RETI, RETX, RETN, or RETE to continue app code Optional nesting of higherpriority interrupts possible See app. note EE192, which covers writing interrupt routines in C (http://www.analog.com/eenotes) February 11, 2012 Veton Kpuska 187 System Peripheral Interrupts The processor system has numerous peripherals, which therefore require many supporting interrupts. Table 47 lists: The Peripheral Interrupt source The Peripheral Interrupt ID used in the System Interrupt Assignment registers (SIC_IARx). See "System Interrupt Assignment Registers (SIC_IARx)" on page 4 30 of BF533 HRM. The generalpurpose interrupt of the core to which the interrupt maps at reset The Core Interrupt ID used in the System Interrupt Assignment registers (SIC_IARx). See "System Interrupt Assignment Registers (SIC_IARx)" on page 4 30 of BF533 HRM. Veton Kpuska 188 February 11, 2012 Peripheral Interrupt Source at Reset State February 11, 2012 Veton Kpuska 189 Peripheral Interrupt Source at Reset State February 11, 2012 Veton Kpuska 190 Initialization of Interrupts If the default assignments shown in Table 47 are acceptable, then interrupt initialization involves only: 1. 2. 3. Initialization of the core Event Vector Table (EVT) vector address entries. Why is this needed? Initialization of the IMASK register Unmasking the specific peripheral interrupts in SIC_IMASK that the system requires February 11, 2012 Veton Kpuska 191 Initialization of Core Vector Table The Event Vector Table (EVT) is a hardware table with sixteen entries that are each 32 bits wide. The EVT contains an entry for each possible core event. Entries are accessed as MMRs, and each entry can be programmed at reset with the corresponding vector address for the interrupt service routine. When an event occurs, instruction fetch starts at the address location in the EVT entry for that event. The processor architecture allows unique addresses to be programmed into each of the interrupt vectors; that is, interrupt vectors are not determined by a fixed offset from an interrupt vector table base address. This approach minimizes latency by not requiring a long jump from the vector table to the actual ISR code. Table 49 lists events by priority. Each event has a corresponding bit in the event state registers: ILAT, IMASK, and IPEND. February 11, 2012 Veton Kpuska 192 Core Event Vector Table February 11, 2012 Veton Kpuska 193 Core Event Vector Table February 11, 2012 Veton Kpuska 194 SPORT Interrupts Initialization Init_Sport_Interrupts() Initialize.c : Init_Sport_Interrupts() //--------------------------------------------------------------------------/ / // Function: Init_Interrupts // // // // Description: Initialize Interrupt for Sport0 RX // //--------------------------------------------------------------------------/ / void Init_Sport_Interrupts(void) { // Set Sport0 RX (DMA1) interrupt priority to 2 = IVG9 *pSIC_IAR0 = 0xffffffff; *pSIC_IAR1 = 0xffffff2f; *pSIC_IAR2 = 0xffffffff; // assign ISRs to interrupt vectors // Sport0 RX ISR -> IVG 9 register_handler(ik_ivg9, Sport0_RX_ISR); // enable Sport0 RX interrupt *pSIC_IMASK = 0x00000200; ssync(); } February 11, 2012 Veton Kpuska 196 System Interrupt Assignment Registers (SIC_IARx) The relative priority of peripheral interrupts can be set by mapping the peripheral interrupt to the appropriate generalpurpose interrupt level in the core. The mapping is controlled by the System Interrupt Assignment register settings, as detailed in Figure 49, Figure 410, and Figure 411. If more than one interrupt source is mapped to the same interrupt, they are logically ORed, with no hardware prioritization. Software can prioritize the interrupt processing as required for a particular system application. February 11, 2012 Veton Kpuska 197 Mapping of Values in SIC_IARx Register to General Purpose Interrupt Table 48 defines the value to write in SIC_IARx to configure a peripheral for a particular IVG priority. February 11, 2012 Veton Kpuska 198 System Interrupt Assignment Register *pSIC_IAR0 = 0xffffffff; *pSIC_IAR0 = 0xffffffff; 0xffffffff = 1111 1111 1111 1111 0xffffffff = 1111 1111 1111 1111 1111 1111 1111 1111 1111 1111 1111 1111 All Corresponding Interrupts in All Corresponding Interrupts in this register are set to lowest priority this register are set to lowest priority February 11, 2012 Veton Kpuska 199 System Interrupt Assignment Register *pSIC_IAR1 = 0xffffff2f; *pSIC_IAR1 = 0xffffff2f; 0xffffff2f = 1111 1111 1111 1111 0xffffff2f = 1111 1111 1111 1111 1111 1111 0010 1111 1111 1111 0010 1111 DMA1 (SPORT0 RX) DMA1 (SPORT0 RX) is set to "2" priority is set to "2" priority February 11, 2012 Veton Kpuska 200 System Interrupt Assignment Register *pSIC_IAR2 = 0xffffffff; *pSIC_IAR2 = 0xffffffff; 0xffffffff = 1111 1111 1111 1111 0xffffffff = 1111 1111 1111 1111 1111 1111 1111 1111 1111 1111 1111 1111 Do not forget to set appropriate Do not forget to set appropriate Interrupt Priority Values for PF A Interrupt Priority Values for PF A interrupts. interrupts. February 11, 2012 Veton Kpuska 201 Initialize.c : Init_Sport_Interrupts() //--------------------------------------------------------------------------/ / // Function: Init_Interrupts // // // // Description: Initialize Interrupt for Sport0 RX // //--------------------------------------------------------------------------/ / void Init_Sport_Interrupts(void) { // Set Sport0 RX (DMA1) interrupt priority to 2 = IVG9 *pSIC_IAR0 = 0xffffffff; *pSIC_IAR1 = 0xffffff2f; *pSIC_IAR2 = 0xffffffff; // assign ISRs to interrupt vectors // Sport0 RX ISR -> IVG 9 register_handler(ik_ivg9, Sport0_RX_ISR); // enable Sport0 RX interrupt *pSIC_IMASK = 0x00000200; ssync(); } February 11, 2012 Veton Kpuska 202 Interrupt Handler Support The Blackfin C/C++ compiler provides support for interrupts and other events used by the Blackfin processor architecture (see Table 125 of VisualDSP++ 4.5 C/C++ Compiler and Library Manual for Blackfin Processors). The Blackfin system has several different classes of events, not all of which are supported by the ccblkfn compiler. Handlers for these events are called Interrupt Service Routines (ISRs). The compiler provides facilities for defining an ISR function, registering it as an event handler, and for obtaining the saved processor context. February 11, 2012 Veton Kpuska 203 Defining an ISR To define a function as an ISR, the sys/exception.h header must be included and the function must be declared and defined using macros defined within this header file. Where is this file included in the LED Lab #1 example? There is a macro for each of the three kinds of events the compiler supports: EX_INTERRUPT_HANDLER EX_EXCEPTION_HANDLER EX_NMI_HANDLER Which macro was used in Lab #1 exercise? By default, the ISRs generated by the compiler are not reentrant; they disable the interrupt system on entry, and reenable it on exit. You may also define ISRs for interrupts which are reentrant, and which reenable the interrupt system soon after entering the ISR. Veton Kpuska 204 February 11, 2012 Example #include <sys/exception.h> static int number_of_interrupts; EX_INTERRUPT_HANDLER(my_isr) { number_of_interrupts++; } This example declares and defines my_isr() as a handler for interrupttype events. The macro used for defining the ISR is also suitable for declaring it as a prototype: EX_INTERRUPT_HANDLER(my_isr); Example: Talkthrough.h: EX_INTERRUPT_HANDLER(Sport0_RX_ISR); February 11, 2012 Veton Kpuska 205 Registering an ISR ISRs, once defined, can be registered in the Event Vector Table (EVT) using the register_handler_ex() function or the register_handler() function, both of which also update the IMASK register so that interrupt can take effect. Only the register_handler_ex() function will be discussed here, as it is an extended version of the register_handler() function. Refer to "register_handler_ex" on page 3236 of VisualDSP++ 4.5 C/C++ Compiler and Library Manual for Blackfin Processors for more information about it. The register_handler_ex() function takes three parameters, defining the event, the ISR, and whether the interrupt should be enabled, disabled, or left in its current state. It also returns the previously registered ISR (if any). The event is specified using the interrupt_kind enumeration from exception.h. February 11, 2012 Veton Kpuska 206 Registering an ISR typedef enum { ik_emulation, ik_reset, ik_nmi, ik_exception, ik_global_int_enable, ik_hardware_err, ik_timer, ik_ivg7, ik_ivg8, ik_ivg9, ik_ivg10, ik_ivg11, ik_ivg12, ik_ivg13, ik_ivg14, ik_ivg15 } interrupt_kind; ex_handler_fn register_handler_ex(interrupt_kind kind, ex_handler_fn fn, int enable); Which enumeration types were used in the LAB exercise? February 11, 2012 Veton Kpuska 207 register_handler Register event handlers Synopsis #include <sys/exception.h> ex_handler_fn register_handler (interrupt_kind kind, ex_handler_fn fn); Description The register_handler function determines how the hardware event kind is handled. This is done by registering the function pointed to by fn as a handler for the event and updating the IMASK register so that interrupt can take effect. The kind event is an enumeration identifying each of the hardware events--interrupts and exceptions--accepted by the Blackfin processor. The register_handler_ex function provides an extended and more functional interface than register_handler. Note: For the values for interrupt_kind, refer to "Registering an ISR" on page 1251 of VisualDSP++ 4.5 C/C++ Compiler and Library Manual for Blackfin Processors presented also in the previous slide . The fn must be one of the values are listed in the table in the next slide. February 11, 2012 Veton Kpuska 208 Table of fn values Which function value was used in the Lab exercise? February 11, 2012 Veton Kpuska 209 Initialize.c : Init_Sport_Interrupts() //--------------------------------------------------------------------------/ / // Function: Init_Interrupts // // // // Description: Initialize Interrupt for Sport0 RX // //--------------------------------------------------------------------------/ / void Init_Sport_Interrupts(void) { // Set Sport0 RX (DMA1) interrupt priority to 2 = IVG9 *pSIC_IAR0 = 0xffffffff; *pSIC_IAR1 = 0xffffff2f; *pSIC_IAR2 = 0xffffffff; // assign ISRs to interrupt vectors // Sport0 RX ISR -> IVG 9 register_handler(ik_ivg9, Sport0_RX_ISR); // enable Sport0 RX interrupt *pSIC_IMASK = 0x00000200; ssync(); } Unmasking of the Interrupt by setting Unmasking of the Interrupt by setting corresponding bits in SIC_IMASK register. corresponding bits in SIC_IMASK register. February 11, 2012 Veton Kpuska 210 SIC_IMASK Register The System Interrupt Mask register (SIC_IMASK, shown in Figure 48 Blacfin BF533 HRM) allows masking of any peripheral interrupt source at the System Interrupt Controller (SIC), independently of whether it is enabled at the peripheral itself. A reset forces the contents of SIC_IMASK to all 0s to mask off all peripheral interrupts. Writing a 1 to a bit location turns off the mask and enables the interrupt. Although this register can be read from or written to at any time (in Supervisor mode), it should be configured in the reset initialization sequence before enabling interrupts. February 11, 2012 Veton Kpuska 211 SIC_IMASK Register *pSIC_IMASK = 0x00000200; *pSIC_IMASK = 0x00000200; 0x00090000 = 0000 0000 0000 0001 0x00090000 = 0000 0000 0000 0001 0000 0010 0000 0000 0000 0010 0000 0000 DMA Interrupt DMA Interrupt SPORT0 RX SPORT0 RX Do not forget to Do not forget to Unmask Interrupt Unmask Interrupt associated with associated with PF PF Interrupt A Interrupt A February 11, 2012 Veton Kpuska 212 Synchronization Synchronization Functions void csync(void) void ssync(void) These two functions provide synchronization. The csync() function is a coreonly synchronization--it flushes the pipeline and store buffers. The ssync() function is a system synchronization, and also waits for an ACK instruction from the system bus. February 11, 2012 Veton Kpuska 213 Servicing Interrupts SPORT0_ISR SPORT0_ISR //--------------------------------------------------------------------------// // Function:Sport0_RX_ISR // // // // Description: This ISR is executed after a complete frame of input data // // has been received. The new samples are stored in // // iChannel0LeftIn, iChannel0RightIn, iChannel1LeftIn and // // iChannel1RightIn respectively. Then the function // // Process_Data() is called in which user code can be executed. // // After that the processed values are copied from the // // variables iChannel0LeftOut, iChannel0RightOut, // // iChannel1LeftOut and iChannel1RightOut into the dma // // transmit buffer. // //--------------------------------------------------------------------------// EX_INTERRUPT_HANDLER(Sport0_RX_ISR) { // confirm interrupt handling *pDMA1_IRQ_STATUS = 0x0001; // copy input data from dma input buffer into variables iChannel0LeftIn = iRxBuffer1[INTERNAL_ADC_L0]; iChannel0RightIn = iRxBuffer1[INTERNAL_ADC_R0]; iChannel1LeftIn = iRxBuffer1[INTERNAL_ADC_L1]; iChannel1RightIn = iRxBuffer1[INTERNAL_ADC_R1]; // call function that contains user code Process_Data(); February 11, 2012 Veton Kpuska 215 DMA Interrupt Status Registers February 11, 2012 Veton Kpuska 216 DMAx_IRQ_STATUS/MDMA_yy_IRQ_ STATUS Register DMAx_IRQ_STATUS/MDMA_yy_IRQ_STATUS Register The Interrupt Status register (DMAx_IRQ_STATUS/MDMA_yy_IRQ_STATUS), shown in Figure 913, contains bits that record whether the DMA channel: Is enabled and operating, enabled but stopped, or disabled. Is fetching data or a DMA descriptor. Has detected that a global DMA interrupt or a channel interrupt is being asserted. Has logged occurrence of a DMA error. Note the DMA_DONE interrupt is asserted when the last memory access (read or write) has completed. Note: For a memory transfer to a peripheral, there may be up to four data words in the channel's DMA FIFO when the interrupt occurs. At this point, it is normal to immediately start the next work unit. If, however, the application needs to know when the final data item is actually transferred to the peripheral, the application can test or poll the DMA_RUN bit. As long as there is undelivered transmit data in the FIFO, the DMA_RUN bit is 1. February 11, 2012 Veton Kpuska 217 DMAx_IRQ_STATUS Register *pDMA1_IRQ_STATUS = 0x0001; *pDMA1_IRQ_STATUS = 0x0001; February 11, 2012 Veton Kpuska 218 SPORT0_ISR //--------------------------------------------------------------------------// // Function:Sport0_RX_ISR // // // // Description: This ISR is executed after a complete frame of input data // // has been received. The new samples are stored in // // iChannel0LeftIn, iChannel0RightIn, iChannel1LeftIn and // // iChannel1RightIn respectively. Then the function // // Process_Data() is called in which user code can be executed. // // After that the processed values are copied from the // // variables iChannel0LeftOut, iChannel0RightOut, // // iChannel1LeftOut and iChannel1RightOut into the dma // // transmit buffer. // //--------------------------------------------------------------------------// EX_INTERRUPT_HANDLER(Sport0_RX_ISR) { // confirm interrupt handling *pDMA1_IRQ_STATUS = 0x0001; // copy input data from dma input buffer into variables iChannel0LeftIn = iRxBuffer1[INTERNAL_ADC_L0]; iChannel0RightIn = iRxBuffer1[INTERNAL_ADC_R0]; iChannel1LeftIn = iRxBuffer1[INTERNAL_ADC_L1]; iChannel1RightIn = iRxBuffer1[INTERNAL_ADC_R1]; // call function that contains user code Process_Data(); February 11, 2012 Veton Kpuska 219 SPORT0_ISR // copy processed data from variables into dma output buffer iTxBuffer1[INTERNAL_DAC_L0] = iChannel0LeftOut; iTxBuffer1[INTERNAL_DAC_R0] = iChannel0RightOut; iTxBuffer1[INTERNAL_DAC_L1] = iChannel1LeftOut; iTxBuffer1[INTERNAL_DAC_L2] = iChannel1LeftOut; iTxBuffer1[INTERNAL_DAC_R1] = iChannel1RightOut; iTxBuffer1[INTERNAL_DAC_R2] = iChannel1RightOut; } February 11, 2012 Veton Kpuska 220 Processing Input Audio Data Process_data() Process_data() #include "Talkthrough.h" #include <btc.h> //// // Function: Process_Data() // // Description: This function is called from inside the SPORT0 ISR every // time a complete audio frame has been received. The new // input samples can be found in the variables iChannel0LeftIn, // iChannel0RightIn, iChannel1LeftIn and iChannel1RightIn // respectively. The processed data should be stored in // iChannel0LeftOut, iChannel0RightOut, iChannel1LeftOut, // iChannel1RightOut, iChannel2LeftOut and iChannel2RightOut // respectively. //// void Process_Data(void) { iChannel1LeftOut = iChannel1LeftIn; iChannel1RightOut = iChannel1RightIn; //write data to BTC channel int nValue; extern int BTCLeftVolume; extern int BTCRightVolume; // // // // // // // // // // February 11, 2012 Veton Kpuska 222 Process_data() nValue = (iChannel0LeftIn >> 8); //upper 24bits nValue = (nValue >> BTCLeftVolume); //volume btc_write_value(0, (unsigned int*)&nValue, sizeof(nValue)); iChannel0LeftOut = (nValue << 8); nValue = (iChannel0RightIn >> 8); //upper 24bits nValue = (nValue >> BTCRightVolume); //volume btc_write_value(1, (unsigned int*)&nValue, sizeof(nValue)); iChannel0RightOut = (nValue << 8); } February 11, 2012 Veton Kpuska 223 ...
View Full Document

This note was uploaded on 02/10/2012 for the course ECE 3551 taught by Professor Staff during the Spring '11 term at FIT.

Ask a homework question - tutors are online