Freescale Semiconductor Application Note AN2865 Rev. 4, 04/2010 Qorivva Simple Cookbook “Hello World” Programs to Exercise Common Features on MPC5500 & MPC5600 Microcontrollers by: Steve Mihalik Field Applications This application note contains software examples to use when getting started using MPC5500 and MPC5600 family devices. Complete code is available for downloading to target such as an evaluation board. Table 1. MPC5500 & MPC5600 Family Definitions for this Application Note Family Name Devices Included MPC551x MPC5514, MPC5515, MPC5516, MPC5517 MPC555x MPC5533, MPC5534, MPC5553, MPC5554, MPC5561, MPC5565, MPC5566, MPC5567, MPC5632M, MPC5633M, MPC5634M MPC56xxB/P/S MPC5602B or C, MPC5603B, MPC5604 B or C, MPC5604P, MPC5602S, MPC5604S, MPC5606S The family definitions in Table 1 are used to categorize the different devices discussed in this document. Examples included here will illustrate any differences between families, and between the different members in a family. © Freescale Semiconductor, Inc., 2005–2010. All rights reserved. Contents 1 Time Base: Time Measurement . . . . . . . . . . . . . . . . . . . 5 2 Interrupts: Decrementer. . . . . . . . . . . . . . . . . . . . . . . . . . 7 3 Interrupts: Fixed-Interval Timer . . . . . . . . . . . . . . . . . . . 14 4 INTC: Software Vector Mode. . . . . . . . . . . . . . . . . . . . . 22 5 INTC: Hardware Vector Mode . . . . . . . . . . . . . . . . . . . . 35 6 INTC: Software Vector Mode, VLE Instructions . . . . . . 48 7 INTC: Hardware Vector Mode, VLE Instructions . . . . . . 66 8 MMU: Create TLB Entry . . . . . . . . . . . . . . . . . . . . . . . . 84 9 Cache: Cache as RAM . . . . . . . . . . . . . . . . . . . . . . . . . 87 10 PLL: Initializing System Clock (MPC551x, MPC55xx). . 91 11 PLL: Initializing System Clock (MPC56xxB/P/S) . . . . . . 98 12 FMPLL: Frequency Modulation . . . . . . . . . . . . . . . . . . 107 13 Modes: Low Power (MPC56xxB/S) . . . . . . . . . . . . . . . 110 14 eDMA: Block Move . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 15 eSCI: Simple Transmit and Receive . . . . . . . . . . . . . . 130 16 eSCI: LIN Transmit . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 17 LINFlex: LIN Transmit . . . . . . . . . . . . . . . . . . . . . . . . . 139 18 eMIOS: Modulus Counter, OPWM Functions . . . . . . . 147 19 eMIOS: PEC, OPWFM Functions . . . . . . . . . . . . . . . . 156 20 eTPU: Set 1 PWM Function . . . . . . . . . . . . . . . . . . . . 162 21 eQADC: Single Software Scan . . . . . . . . . . . . . . . . . . 168 22 ADC: Software Trigger, Continuous Scan . . . . . . . . . . 171 23 ADC - CTU: eMIOS Trigger (MPC560xB) . . . . . . . . . . 178 24 DSPI: SPI to SPI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 25 FlexCAN Transmit and Receive . . . . . . . . . . . . . . . . . 201 26 Flash: Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . 220 Appendix AInterrupt Alignment Summary . . . . . . . . . . . . . . 229 Appendix BSingle Core Build Files . . . . . . . . . . . . . . . . . . . 230 Appendix CMPC56xxB/P/S Peripheral Clocks . . . . . . . . . . 246 Source code, necessary build files, and debugger scripts are available for all examples. Example code provides two executable outputs: one that locates the program in internal flash, and another that locates the program in internal RAM. For RAM-based programs, debugger scripts are provided to initialize the MMU, initialize SRAM, and set the instruction pointer to the beginning of main code. For flash-based programs: • The MMU is initialized by Boot Assist Module (BAM) code that generally executes after reset. • SRAM is initialized by added code to a modified startup (“crt0”) file. • The instruction pointer (as well as some reset configuration settings) is defined by the Reset Configuration Half Word (RCHW) in the boot section of the startup file for flash. NOTE For EVBs with 40 MHz Crystal: MPC5561 and MPC5567 EVBs use 40 MHz crystals instead of the usual 8 MHz crystal. In this case, the default frequency will be 30 MHz instead of 12 MHz. Examples that alter the PLL frequency will need software modification for proper operation using a 40 MHz crystal. (For more information, see Section 10.2.2, “MPC555x,” part of Section 8.2, “Design,” in Section 8, “PLL: Initializing System Clock,” for code changes when using 40 MHz crystal.) Qorivva Simple Cookbook, Rev. 4 2 Freescale Semiconductor Table 2. Devices and Applicable Examples MPC551x Devices MPC5514, MPC5515, MPC5516, MPC5517 Example MPC555x Devices MPC56xxBPS Devices MPC5533, MPC5553 MPC5632M, MPC560xB, MPC5534 MPC5554, MPC5633M, MPC560xP, MPC5561, MPC5634M MPC560xS MPC5565, MPC5566, MPC5567 1 Time Base: Time Measurement x x x x 2 Interrupts: Decrementer x x x x 3 Interrupts: Fixed-Interval Timer x x x x 4 INTC: Software Vector Mode x x x x 5 INTC: Hardware Vector Mode x x x x 6 INTC: Software Vector Mode, VLE x x x 7 INTC: Hardware Vector Mode, VLE x x x 8 MMU: Create TLB Entry x 9 Cache: Cache as RAM x x x x 10 PLL: Initializing System Clock x x x x 11 PLL: Initializing System Clock x 12 FMPLL: Frequency Modulation x 13 Modes: Low Power x 14 eDMA: Block Move x x x x 15 eSCI: SImple Transmit & Receive x x x x 16 eSCI: LIN x x x x 17 LINFlex: LIN Transmit x x 18 eMIOS: Modulus Counter, OPWM x x x 19 eMIOS: PEC, OPWFM x x (except MPC56xxP) x 20 eTPU: Set 1 PWM Function 21 eQADC: Single Software Scan x x x (except MPC5561) x x x x 22 ADC: SW Trigger, Cont Scan x 23 ADC-CTU: eMIOS Trigger MPC560xB 24 DSPI: SPI to SPI x x x x x 25 FlexCAN: Transmit and Receive x x x x x 26 Flash: Configuration x x x x x Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 3 Revision History Revision 2 (Earlier revisions not tracked) Minor text changes in various examples, plus code modifications for example 2, “Interrupts: Decrementer,” and example 19, “Flash: Configuration.” 11/2008 Revision 3 • Added 3 new examples: PLL: System Clock (MPC56xxB/P/S), INTC: SW Vector Mode, VLE Instructions, INTC: HW Vector Mode, VLE Instructions • Expanded and updated Flash: Configuration • Updated 5 other examples to support MPC560xB, MPC560xP, MPC560xS (“MPC56xxB/P/S”) • Revised FlexCAN example • Minor text changes in other examples 07/2009 Revision 4 • Added 4 new examples: Modes: Low Power LINFlex: LIN Transmit ADC: Software trigger, Continuous Scan CTU: eMIOS Trigger • Updated other MPC56xxB/P/S examples • Minor text changes in other examples 04/2010 The revision number was not incremented because no substantive changes were made to the content. • Front page: Add SafeAssure branding. • Title and first page: Add Qorivva branding. • Back page: Apply new back page format. 04/2012 Qorivva Simple Cookbook, Rev. 4 4 Freescale Semiconductor 1 Time Base: Time Measurement 1.1 Description Task: Using the Time Base (TB), measure the number of system clocks to execute a section of code in any type of memory. Put the result into a register. This program demonstrates how to manage special purpose registers (spr’s). Special purpose registers are not memory-mapped, and are read or written using a general purpose register (gpr). Only two instructions are used with spr’s: move to spr (mtspr) and move from spr (mfspr). The 64-bit Time Base consists of two 32-bit special purpose registers: TBL for the lower 32 bits and TBU for the upper 32 bits. It is enabled by setting the Time Base Enable (TBE) bit in the Hardware Implementation Register 0 (spr HID0). When enabled it counts at system clock frequency, after some initial delay due to pipelining. If sysclk = 12 MHz (the default for MPC555x devices with an 8 MHz crystal), it would take about six minutes for TBL to overflow to TBU. MPC551x devices have a default frequency of 16 MHz. NOTE Debuggers may or may not stop the time base when code is not running, such as at a breakpoint. MPC5500 Time Base Crystal 8 MHz Clocks and PLL sysclk spr TBU spr TBL Enable spr HID0 TBE Figure 1. Time Base Example Exercise: Modify the program to record the TBL and TBU at the start of the desired code sequence, and move their values to spr SPRG0 and spr SPRG1. At the end of the measured code, record both again and move their values to spr SPRG2 and spr SPRG3. Examine these registers and calculate the time difference. Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 5 1.2 Design Table 3. Time Base: Time Measurement Step Code 1 Initialize Time Base = 0 li mttbu mttbl r4, 0 r4 r4 # Load immediate data of 0 to r4 # Move r4 to TBU # Move r4 to TBL 2 Enable Time Base: set HID0[TBE]=1 mfhid0 li or mthid0 r5 r4, 0x4000 r5, r4, r5 r5 # Move from spr HID0 to r5 # Load immediate data of 0x4000 to r4 # OR r4 (0x4000 0000) with r4 (HID0 value) # Move result to HID0 3 Execute some code nop nop nop nop nop 4 Record TBL mftbl r5 # Move TBL to r5 to store TBL value 1.3 Code li mttbu mttbl r4, 0 r4 r4 mfhid0 li or mthid0 r5 r4, 0x4000 r5, r4, r5 r5 nop nop nop nop nop mftbl r5 # # # # # # # # # # INITIALIZE TIME BASE=0 Load immediate data of 0 to r4 Move r4 to TBU Move r4 to TBL ENABLE TIME BASE Move from spr HID0 to r5 (copies HID0) Load immed. data of 0x4000 to r4 OR r4 (0x0000 4000) with r5 (HID0 value) Move result to HID0 EXECUTE SOME CODE # RECORD TBL # Move TBL to r5 to store TBL value Qorivva Simple Cookbook, Rev. 4 6 Freescale Semiconductor 2 Interrupts: Decrementer 2.1 Description Task: Use the Decrementer (DEC) exception to generate a periodic interrupt roughly every second. Design the interrupt handler, including interrupt service routine (ISR), to be written entirely in assembler without enabling nested interrupts. The ISR simply counts the number of Decrementer interrupts, toggles a GPIO output, and clears the Decrementer interrupt flag. Assume the system clock runs at the default frequency of 16 MHz for MPC551x (16 MHz Internal Reference Clock) or 12 MHz for MPC555x (1.5 8 MHz crystal). Exercise: Connect the output to an LED on an MPC55xx evaluation board or oscilloscope. After verifying proper operation, change the Decrementer timeout to 1 kHz. MPC5500 Crystal 8 MHz Clocks and PLL Decrementer Automatic Reload (spr DECAR) Default sysclk: 16 MHz (MPC551x) 12 MHz (MPC555x) 8 MHz (MPC563x) Decrementer (spr DEC) Enable GPIO spr HID0 TBE DEC interrupt request to CPU core caused by the Decrementer passing through zero Figure 2. Decrementer Example Table 4. Signals for Decrementer Example MPC551x Family Signal Pin Name SIU PCR No. GPIO PH2 114 MPC555x Family Package Pin No. 144 QFP 176 QFP 208 BGA 22 30 L1 Function Name SIU PCR No. GPIO[114] 114 Package Pin No. EVB 496 BGA 416 BGA 324 BGA 208 BGA 144 QFP xPC 563M M5 N3 L3 N3 52 PJ9–1 Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 7 2.2 Design For MPC551x, the DEC interrupt causes the CPU to vector to the sum of the common Interrupt Vector Prefix (IVPR0:19) plus a fixed offset for each interrupt, which is 16 bytes times the IVOR number. The MPC551x program flow is shown below. main.c file main { init IVPR init Decrementer init GPIO enable interrupts wait forever } Decrementer interrupt taken vector to IVPR0:19 + (10 16) bytes ivor_branch_table.s file handlers.s file IVOR Branch Table b IVOR0trap b IVOR1trap ... b IVOR10Handler ... IVOR10Handler: prologue (save registers) execute DecISR (increment ctr, toggle pin, clear flag) epilogue (restore registers, return) Figure 3. MPC551x Program Flow For MPC555x, the DEC interrupt causes the CPU to vector to the sum of the common Interrupt Vector Prefix (IVPR0:15) plus the interrupt’s individual Interrupt Vector Offset (IVOR1016:27). The MPC555x program flow is shown below, followed by design steps and a stack frame implementation. main.c file main { init IVPR, IVOR10 init Decrementer init GPIO enable interrupts wait forever } Decrementer interrupt taken vector to IVPR0:15 + IVOR1016:27 handlers.s file IVOR10Handler: prologue (save registers) execute DecISR (increment ctr, toggle pin, clear flag) epilogue (restore registers, return) Figure 4. MPC555x Program Flow Qorivva Simple Cookbook, Rev. 4 8 Freescale Semiconductor Table 5. Initialization Steps Relevant Bit Fields Step Pseudo Code MPC551x Global variable Counter for decrementer initIrqVectors int DECctr = 0 Load the common interrupt prefix to IVPR. Prefix value is defined in the link file. spr IVPR = __IVPR_VALUE MPC555x: Initialize decrementer interrupt offset to the lower half of IVOR10 handler’s address. (MPC551x note: Each core has its own spr IVPR.) initDEC MPC555x – Load initial DEC value of 12M = 0xB71B00. This provides 1 sec timeout for 12 MHz sysclk, and 0.75 sec timeout for 16 MHz sysclk. spr IVOR10 = IVOR10Handler@l spr DEC = 0x00B7 1B00 Load decrementer automatic reload value of 12M. spr DECAR = 0x00B7 1B00 Enable decrementer interrupt. DIE = 1 Enable decrementer automatic reload on timeout. ARE = 1 spr TCR = 0x0000 0440 Start DEC (and Time Base) counting. spr HID0 = 0x0000 4000 TBEN = 1 initGPIO Initialize GPIO as output: Pad assignment is GPIO (default). PA = 0 (default) Output buffer is enabled. OBE = 1 Input buffer is also enabled (allows monitoring pad). IBE = 1 SIU_PCR[114] = 0x0303 enableExtIrq Enable external interrupts by setting MSR[EE] = 1. (Enables requests from INTC, DEC, FIT to be recognized if enabled and pending.) wrteei 1 waitloop wait forever while 1 Because the program’s interrupt handler is entirely in assembler, only the registers that are used will need to be saved in the stack. Two registers are used in the interrupt service routine, so only these two registers are saved and restored in the prologue and epilogue, besides the stack pointer (r1). External input interrupts, enabled by MSR[EE], are not re-enabled in the interrupt handler, so SRR0:1 need not be saved in the stack frame. Table 6. Stack Frame for Decrementer Interrupt (16 bytes) Stack Frame Area 32-bit GPRs Register Location in Stack r4 sp + 0x0C r3 sp + 0x08 LR area LR placeholder sp + 0x04 back chain r1 (which is sp) sp Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 9 Table 7. IVOR10 (Decrementer) Interrupt Handler Steps (MPC555x: Must be 16 byte aligned and within 64KB of address in IVPR. MPC551x does not require alignment because it is the destination of a branch instruction.) Relevant Bit Fields Step prolog Create stack frame stwu sp, –0x10 (sp) Save registers in stack frame DecISR store required registers Increment DECctr DECctr++ Toggle GPIO SIU_GPDO[114] = SIU_GPDI[114]++ Clear decrementer interrupt flag epilog Pseudo Code (MPC551x & MPC555x) DIS = 1 spr TSR = 0x0800 0000 Restore registers from stack frame restore registers Restore stack frame space addi sp, sp, 0x50 Return rfi Qorivva Simple Cookbook, Rev. 4 10 Freescale Semiconductor 2.3 2.3.1 /* /* /* /* /* /* /* /* /* Code main.c file main.c - Decrementer interrupt example */ Rev 1.0 April 19, 2004 S.Mihalik, Copyright Freescale, 2007 All Rights Reserved */ Rev 1.1 May 15 2006 SM- Changed GPIO205 to GPIO195 for use in 324, 208 packages */ Rev 1.2 Aug 12 2006 SM- Made variable i volatile */ Rev 1.3 Jun 13 2007 SM- Passed spr IVPR value from link file, changed GPIO pin */ and for MPC551x removed code to load spr IVOR10 */ Notes: */ 1. MMU not initialized; must be done by debug scripts or BAM */ 2. SRAM not initialized; must be done by debug scripts or in a crt0 type file */ #include "mpc563m.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ extern IVOR10Handler(); extern uint32_t __IVPR_VALUE; /* Interrupt Vector Prefix value from link file */ uint32_t DECctr = 0; /* Counter for Decrementer interrupts */ vuint32_t GPDI_114_ADDR = (vuint32_t)&SIU.GPDI[114].R; /* GPDI[114] reg. addr. */ vuint32_t GPDO_114_ADDR = (vuint32_t)&SIU.GPDO[114].R; /* GPDO[114] reg. addr. */ asm void initIrqVectors(void) { lis r0, __IVPR_VALUE@h /* IVPR value is passed from link file */ ori r0, r0,__IVPR_VALUE@l /* Note: IVPR lower bits are unused in MPC555x*/ mtivpr r0 /* The following two lines are required for MPC555x, and are not used for MPC551x*/ li r0, IVOR10Handler@l /* IVOR10(Dec) = lower half of handler address */ mtivor10 r0 } asm void initDEC(void) { lis r0, 0x00B7 ori r0, r0, 0x1B00 mtdec r0 mtdecar r0 lis r0, 0x0440 mttcr r0 li r0, 0x4000 mthid0 r0 } void main (void) { volatile uint32_t i=0; initIrqVectors(); initDEC(); SIU.PCR[114].R = 0x0303; asm (" wrteei 1"); while (1) { i++; } } /* Load initial DEC value of 12M (0x00B7 1B00) */ /* Load same initial value to DECAR */ /* Enable DEC interrupt and auto-reload*/ /* Enable Time Base and Decrementer (set TBEN) */ /* /* /* /* /* /* Dummy idle counter */ Initialize interrupt vectors registers*/ Initialize Decrementer routine */ Init. pin for GPIO output that can be read as input */ Enable external interrupts (INTC, DEC, FIT) */ Loop forever */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 11 2.3.2 handler.s file # handlers.s - Decrementer (IVOR10) interrupt example # Rev 1.0: Sept 2, 2004, S Mihalik, # Rev 1.1: May 15, 2006 S.M.- Use GPIO[195] instead of GPIO[205] # Rev 1.2: Jun 13, 2007 S.M. Added .section .ivor_handlers for linking # and changed GPIO address to global variable # Rev 1.3 Nov 1 2008 SM - Used r4 instead of r0 to clear DIS flag in spr TSR# Copyright Freescale Semiconductor, Inc. 2007. All rights reserved # STACK FRAME DESIGN: Depth: 4 words (0x10, or 16 bytes) # ************* ______________ # 0x0C * r4 * |GPR Save Area # 0x08 * r3 * ___|__________ # 0x04 * resvd- LR * Reserved for calling function # 0x00 * SP(GPR1) * Backchain (same as gpr1 in GPRs) # ************* .globl IVOR10Handler .extern DECctr .extern GPDI_114_ADDR .extern GPDO_114_ADDR # Counter of Decrementer interrupts taken # GPDI[114] register address # GPDO[114] register address .section .ivor_handlers # # .align 4 .align 16 # Align IVOR handlers on a 16 byte boundary for MPC555x # GHS, Cygnus, Diab(default) use .align 4 # CodeWarrior requires .align 16 IVOR10Handler: prolog: stwu stw stw r1, -0x10 (r1) r3, 0x08 (r1) r4, 0x0C (r1) DECisr: lis lwz addi stw r3, r4, r4, r4, lis lwz lbz addi lis lwz stb lis mtspr epilog: lwz lwz addi rfi DECctr@ha DECctr@l (r3) r4, 0x1 DECctr@l (r3) # PROLOGUE # Create stack frame and store back chain # Read DECctr # Increment DECctr # Write back new DECctr r3, GPDI_114_ADDR@ha # Get pointer to memory containing GPDI[114] address r3, GPDI_114_ADDR@l(r3) r4, 0 (r3) # Read pin input state from register GPDI[114] r4, r4, 1 # Add one to state for toggle effect r3, GPDO_114_ADDR@ha # Get pointer to memory containing GPDO[114] address r3, GPDO_114_ADDR@l(r3) r4, 0 (r3) # Output toggled GPIO 114 pin state to reg. GPDO[114] r4, 0x0800 336, r4 r4, 0x0C (r1) r3, 0x08 (r1) r1, r1, 0x10 # Write "1" clear Dec Interrupt Status (DIS) flag # DIS flag is in spr TSR (spr 336) # EPILOGUE # Restore gprs # Restore space on stack # End of Interrupt Qorivva Simple Cookbook, Rev. 4 12 Freescale Semiconductor 2.3.3 # # # # # ivor_branch_table.s file (MPC551x only) ivor_branch_table.s - for use with MPC551x only Description: Branch table for 16 MPC551x core interrupts Copyright Freescale 2007. All Rights Reserved Rev 1.0 Jul 6 2007 S Mihalik - Initial version Rev 1.1 Aug 30 2007 SM - Made IVOR10Handler extern .extern IVOR10Handler .section .ivor_branch_table .equ SIXTEEN_BYTES, 16 # 16 byte alignment required for table entries # Diab compiler uses value of 4 (2**4=16) # CodeWarrior, GHS, Cygnus use 16 .align SIXTEEN_BYTES b IVOR0trap # IVOR 0 interrupt handler .align SIXTEEN_BYTES IVOR1trap: b IVOR1trap # IVOR 1 interrupt handler .align SIXTEEN_BYTES IVOR2trap: b IVOR2trap # IVOR 2 interrupt handler .align SIXTEEN_BYTES IVOR3trap: b IVOR3trap # IVOR 3 interrupt handler .align SIXTEEN_BYTES IVOR4trap: b IVOR4trap # IVOR 4 interrupt handler .align SIXTEEN_BYTES IVOR5trap: b IVOR5trap # IVOR 5 interrupt handler .align SIXTEEN_BYTES IVOR6trap: b IVOR6trap # IVOR 6 interrupt handler .align SIXTEEN_BYTES IVOR7trap: b IVOR7trap # IVOR 7 interrupt handler .align SIXTEEN_BYTES IVOR8trap: b IVOR8trap # IVOR 8 interrupt handler .align SIXTEEN_BYTES IVOR9trap: b IVOR9trap # IVOR 9 interrupt handler .align SIXTEEN_BYTES b IVOR10Handler # IVOR 10 interrupt handler (Decrementer) .align SIXTEEN_BYTES IVOR11trap: b IVOR11trap # IVOR 11 interrupt handler .align SIXTEEN_BYTES IVOR12trap: b IVOR12trap # IVOR 12 interrupt handler .align SIXTEEN_BYTES IVOR13trap: b IVOR13trap # IVOR 13 interrupt handler .align SIXTEEN_BYTES IVOR14trap: b IVOR14trap # IVOR 14 interrupt handler .align SIXTEEN_BYTES IVOR15trap: b IVOR15trap # IVOR15 interrupt handler IVOR0trap: Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 13 3 Interrupts: Fixed-Interval Timer 3.1 Description Task: Use the Fixed-Interval Timer (FIT) exception to generate a periodic interrupt roughly every second. Design the FIT’s interrupt handler to allow the interrupt service routine (ISR) to be written in C. The FIT’s ISR simply counts interrupts and toggles a GPIO output. The interrupt handler will not enable nested interrupts. Assume the system clock runs at the default frequency, which is 16 MHz for an MPC551x and 12 MHz for an MPC555x with an 8 MHz crystal. The link files, make file, and start up file are the same as for the Decrementer example except for the output file names. Exercise: Connect the GPIO to an LED on an MPC55xx evaluation board or oscilloscope. After verifying proper operation, change the FIT timeout rate to about 4 Hz. MPC5500 Default sysclk: 16 MHz (MPC551x) Clocks and PLL Crystal 8 MHz spr TBL spr TBU 12 MHz (MPC555x) 8 MHz (MPC563x) Enable spr HID0 GPIO TBE FIT interrupt request to CPU core caused by zero to one transition of selected Time Base bit Figure 5. Fixed-Interval Interrupt Example Table 8. Signals for Fixed-Interrupt Example MPC551x Family Signal Pin Name SIU PCR No. GPIO PH2 114 MPC555x Family Package Pin No. 144 QFP 176 QFP 208 BGA 22 30 L1 Function Name SIU PCR No. GPIO[114] 114 Package Pin No. EVB 496 BGA 416 BGA 324 BGA 208 BGA xPC 563M M5 N3 L3 N3 PJ9–1 Qorivva Simple Cookbook, Rev. 4 14 Freescale Semiconductor 3.2 Design For a FIT timeout to occur, the selected Time Base bit must transition from 0 to 1. As the Time Base continues to increment, the bit will transition back to 0 then to 1 again before the next timeout. Therefore for Time Base bit n (n is counted from the least significant bit of TBL), the timeout period is: 2 2n sysclock periods. For one second timeout using a 12 MHz sysclk, 1 second = 2 2n 83 1/3 ns. Solving for integer n, we find when n=22, there is about a 0.7 second timeout. This n = 22 corresponds to TBL bit 31 – n = TBL bit 9. For MPC551x, the FIT interrupt causes the CPU to vector to the sum of the common Interrupt Vector Prefix (IVPR0:19) plus a fixed offset for each interrupt, which is 16 bytes times the IVOR number. The MPC551x program flow is shown below. main.c file Fixed Interval Timer interrupt taken main { init IVPR init Fixed Interval Timer init GPIO enable interrupts wait forever } FitISR { increment counter toggle pin clear FIT interrupt flag } vector to IVPR0:19 + (11 16) bytes ivor_branch_table.s file handlers.s file IVOR Branch Table b IVOR0trap b IVOR1trap ... b IVOR11Handler ... IVOR11Handler: prologue (save registers) branch to FitISR (increment ctr, toggle pin, clear flag) epilogue (restore registers, return) Figure 6. MPC551x Program Flow For MPC555x, the DEC interrupt causes the CPU to vector to the sum of the common Interrupt Vector Prefix (IVPR0:15) plus the interrupt’s individual Interrupt Vector Offset (IVOR1116:27). The MPC555x program flow is shown below, followed by design steps and a stack frame implementation. main.c file Fixed Interval Timer interrupt taken main { init IVPR, IVOR11 init Fixed Interval Timer init GPIO enable interrupts wait forever } FitISR { increment counter toggle pin clear FIT interrupt flag } vector to IVPR0:15 + IVOR1116:27 handlers.s file IVOR11Handler: prologue (save registers) branch to FitISR (increment ctr, toggle pin, clear flag) epilogue (restore registers, return) Figure 7. MPC555x Program Flow Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 15 Table 9. Initialization Steps Pseudo Code Step Relevant Bit Fields MPC551x Global variable Counter for Fixed Interval Timer initIrqVectors Load the common interrupt prefix to IVPR. Prefix value is defined in the link file. int FITctr = 0 spr IVPR = __IVPR_VALUE MPC555x: Initialize decrementer interrupt offset to the lower half of IVOR11 handler’s address. (MPC551x note: Each core has its own spr IVPR.) initFIT Initialize Timer ControlB • FIT Timeout = 0.7 sec (12 MHz sysclk): use TBLb bit 9 • Enable FIT interrupt – spr IVOR11 = IVOR11Handler@l FPEXT= 0xA, FP = 1 FIE = 1 Initialize Time Base = 0 initGPIO MPC555x spr TCR = 0x0181 4000 spr TBL = spr TBU = 0 Start Time Base (and Decrementer) counting TBEN = 1 Initialize GPIO as output: Pad assignment is GPIO (default) Output buffer is enabled Input buffer is also enabled (allows monitoring pad) PA = 0 (default) OBE = 1 IBE = 1 spr HID0 = 0x0000 4000 SIU_PCR[114] = 0x0303 enableExtIrq Enable external interrupts by setting MSR[EE] = 1 (Enables requests from INTC, DEC, FIT to be recognized if enabled and pending) wrteei 1 waitloop wait forever while 1 The ISR will be written in C, so the handler’s prologue will save all the registers the compiler might use, namely the volatile registers as defined in the e500 ABI.1 External input interrupts, enabled by MSR[EE], are not re-enabled in the interrupt handler, so SRR0:1 need not be saved in the stack frame. 1. The SPE’s accumulator is not saved in the MPC555x prologue. It is assumed SPE instructions are not used in this example. Qorivva Simple Cookbook, Rev. 4 16 Freescale Semiconductor Table 10. Stack Frame for FIT Interrupt Handler (80 bytes) Stack Frame Area 32-bit GPRs Register Location in Stack r12 sp + 0x4C r11 sp + 0x48 r10 sp + 0x44 r9 sp + 0x40 r8 sp + 0x3C r7 sp + 0x38 r6 sp + 0x34 r5 sp + 0x30 r4 sp + 0x2C r3 sp + 0x28 r0 sp + 0x24 CR CR sp + 0x20 locals and padding XER sp + 0x1C CTR sp + 0x18 LR sp + 0x14 padding sp + 0x10 padding sp + 0x0C padding sp + 0x08 LR area LR placeholder sp + 0x04 back chain r1 (which is sp) sp Table 11. Interrupt Handler (IVOR11Handler): Fixed-Interval Timer (MPC555x: Must be 16-byte aligned and within 64 KB of address in IVPR. MPC551x does not require alignment because it is the destination of a branch instruction.) Relevant Bit Fields Step prolog Create stack frame stwu sp, –0x50 (sp) Save registers in stack frame FitISR store required registers Increment FITctr FITctr++ Toggle GPIO114 SIU_GPDO[114] = ~SIU_GPDI[114] Clear FIT interrupt status flag epilog Pseudo Code (MPC551x & MPC555x) FIS = 1 spr TSR = 0x0400 0000 Restore registers from stack frame restore registers Restore stack frame space addi sp, sp, 0x50 Return rfi Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 17 3.3 3.3.1 /* /* /* /* /* /* /* /* /* /* Code main.c file main.c - Fixed Interval Timer interrupt example */ Rev 1.0 April 19, 2004 S.Mihalik, Copyright Freescale, 2004 All Rights Reserved */ Rev 1.1 Sept 1, 2004 SM - simplfied, syntax changes */ Rev 1.2 May 15 2006 SM- Changed GPIO205 to GPIO195 for use in 324, 208 packages */ Rev 1.3 Aug 12 2006 SM - Changed variable i to volatile */ Rev 1.4 Jun 18 2007 SM - Passed spr IVPR value from link file, changed GPIO pin */ and f0r MPC551x removed code to load spr IVOR11 */ Notes: */ 1. MMU not initialized; must be done by debug scripts or BAM */ 2. SRAM not initialized; must be done by debug scripts or in crt0 type file */ #include "mpc563m.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ extern IVOR11Handler(); extern uint32_t __IVPR_VALUE; uint32_t FITctr = 0; /* Needed for MPC555x only */ /* Interrupt Vector Prefix value from link file */ /* Counter for Fixed Interval Timer interrupts */ asm void initIrqVectors(void) { lis r0, __IVPR_VALUE@h /* IVPR value is passed from link file */ ori r0, r0, __IVPR_VALUE@l /* Note: IVPR lower bits are unused in MPC555x */ mtivpr r0 /* The following two lines are required for MPC555x, and are not used for MPC551x */ li r0, IVOR11Handler@l /* IVOR11 = lower half of handler address */ mtivor11 r0 } asm void initFIT(void) { li r0, 0 mttbu r0 mttbl r0 lis r0, 0x0181 ori r0, r0, 0x4000 mttcr r0 li r0, 0x4000 mthid0 r0 } /* Initialize time base to 0 /* Enable FIT interrupt and set*/ /* FP=0, FPEXT=A for 0.7 sec timeout */ void main (void) { volatile uint32_t i = 0; } initIrqVectors(); initFIT(); SIU.PCR[114].R= 0x0303; asm (" wrteei 1"); while (1) { i++; } */ /* Enable Time Base and Decrementer (set TBEN) */ /* Dummy idle counter */ /* /* /* /* /* Initialize interrupt vectors registers*/ Initialize FIT routine */ Initialize GPIO as output. */ Enable external interrupts (INTC, DEC, FIT) */ Loop forever */ asm void ClrFitFlag(void) { lis r0, 0x0400 mttsr r0 /* Write "1" clear FIT Interrupt Status flag */ } void FitISR(void) { FITctr++; /* Increment interrupt counter */ SIU.GPDO[114].R = ~SIU.GPDI[114].R; /* Toggle GPIO output */ ClrFitFlag(); /* Clear FIT's flag */ } Qorivva Simple Cookbook, Rev. 4 18 Freescale Semiconductor 3.3.2 # # # # # handlers.file handlers.s - FIT (IVOR11) interrupt example Rev 1.0: April 9, 2004, S Mihalik, Rev 1.1: June 18, 2007 SM Added .section .ivor_handlers for linking Rev 1.2: Aug 23 2007 DF: Made FitISR .extern Copyright Freescale Semiconductor, Inc. 2007. All rights reserved # STACK FRAME DESIGN: Depth: 20 words (0xA0, or 80 bytes) # ************* ______________ # 0x4C * GPR12 * ^ # 0x48 * GPR11 * | # 0x44 * GPR10 * | # 0x40 * GPR9 * | # 0x3C * GPR8 * | # 0x38 * GPR7 * GPRs (32 bit) # 0x34 * GPR6 * | # 0x30 * GPR5 * | # 0x2C * GPR4 * | # 0x28 * GPR3 * | # 0x24 * GPR0 * ___v__________ # 0x20 * CR * __CR__________ # 0x1C * XER * ^ # 0x18 * CTR * | # 0x14 * LR * locals & padding for 16 B alignment # 0x10 * padding * | # 0x0C * padding * | # 0x08 * padding * ___v__________ # 0x04 * resvd- LR * Reserved for calling function # 0x00 * SP * Backchain (same as gpr1 in GPRs) # ************* .extern FitISR .globl IVOR11Handler .section .ivor_handlers .align 16 prolog: stwu stw stw stw stw stw stw stw stw stw stw stw mfCR stw mfXER stw mfCTR stw mfLR stw bl r1, -0x50 r12, 0x4C r11, 0x48 r10, 0x44 r9, 0x40 r8, 0x3C r7, 0x38 r6, 0x34 r5, 0x30 r4, 0x2C r3, 0x28 r0, 0x24 r0 r0, 0x20 r0 r0, 0x1C r0 r0, 0x18 r0 r0, 0x14 FitISR # Align IVOR handlers on a 16 byte (2**4) boundary # GHS, Cygnus, Diab(default) use .align 4; CodeWarrior .align 16 # PROLOGUE (r1) # Create stack frame and store back chain (r1) # Store gprs (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) # Store CR (r1) # Store XER (r1) # Store CTR (r1) # Store LR (r1) # Execute FIT ISR, but return here Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 19 epilog: lwz mtLR lwz mtCTR lwz mtXER lwz mtcrf lwz lwz lwz lwz lwz lwz lwz lwz lwz lwz lwz addi rfi r0, 0x14 (r1) r0 r0, 0x18 (r1) r0 r0, 0x1C (r1) r0 r0, 0x20 (r1) 0xff, r0 r0, 0x24 (r1) r3, 0x28 (r1) r4, 0x2C (r1) r5, 0x30 (r1) r6, 0x34 (r1) r7, 0x38 (r1) r8, 0x3C (r1) r9, 0x40 (r1) r10, 0x44 (r1) r11, 0x48 (r1) r12, 0x4C (r1) r1, r1, 0x50 # EPILOGUE # Restore LR # Restore CTR # Restore XER # Restore CR # Restore gprs # Restore space on stack # End of Interrupt Qorivva Simple Cookbook, Rev. 4 20 Freescale Semiconductor 3.3.3 # # # # # ivor_branch_table.s file (MPC551x only) ivor_branch_table.s - for use with MPC551x only Description: Branch table for 16 MPC551x core interrupts Copyright Freescale 2007. All Rights Reserved Rev 1.0 Jul 6 2007 S Mihalik - Initial version Rev 1.1 Aug 23 2007 DF - Made IVOR11Handler .extern .extern IVOR11Handler .section .ivor_branch_table .equ SIXTEEN_BYTES, 16 # 16 byte alignment required for table entries # Diab compiler uses value of 4 (2**4=16) # CodeWarrior, GHS, Cygnus use 16 .align SIXTEEN_BYTES b IVOR0trap # IVOR 0 interrupt handler .align SIXTEEN_BYTES IVOR1trap: b IVOR1trap # IVOR 1 interrupt handler .align SIXTEEN_BYTES IVOR2trap: b IVOR2trap # IVOR 2 interrupt handler .align SIXTEEN_BYTES IVOR3trap: b IVOR3trap # IVOR 3 interrupt handler .align SIXTEEN_BYTES IVOR4trap: b IVOR4trap # IVOR 4 interrupt handler .align SIXTEEN_BYTES IVOR5trap: b IVOR5trap # IVOR 5 interrupt handler .align SIXTEEN_BYTES IVOR6trap: b IVOR6trap # IVOR 6 interrupt handler .align SIXTEEN_BYTES IVOR7trap: b IVOR7trap # IVOR 7 interrupt handler .align SIXTEEN_BYTES IVOR8trap: b IVOR8trap # IVOR 8 interrupt handler .align SIXTEEN_BYTES IVOR9trap: b IVOR9trap # IVOR 9 interrupt handler .align SIXTEEN_BYTES IVOR10trap: b IVOR10trap # IVOR 10 interrupt handler .align SIXTEEN_BYTES b IVOR11Handler # IVOR 11 interrupt handler (Fixed Interval Timer) .align SIXTEEN_BYTES IVOR12trap: b IVOR12trap # IVOR 12 interrupt handler .align SIXTEEN_BYTES IVOR13trap: b IVOR13trap # IVOR 13 interrupt handler .align SIXTEEN_BYTES IVOR14trap: b IVOR14trap # IVOR 14 interrupt handler .align SIXTEEN_BYTES IVOR15trap: b IVOR15trap # IVOR15 interrupt handler IVOR0trap: Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 21 4 INTC: Software Vector Mode 4.1 Description Task: Using the Interrupt Controller (INTC) in software vector mode, this program provides two interrupts that show nesting. The first interrupt is generated from an eMIOS channel at a 1 kHz rate using the MPC555x default system clock. For the MPC551x, the faster default system clock produces an eMIOS channel timeout rate of approximately 1 kHz 16/12 = 1.33 kHz. The interrupt handler will re-enable interrupts and later, in its interrupt service routine (ISR), invoke a second interrupt every other time. This provides an approximate 1 millisecond task (ISR) from the eMIOS channel and an approximate 2 millisecond task (ISR) from the software interrupt. The software interrupt will have a higher priority, so the 1 eMIOS ISR is preempted by the software interrupt. Both ISRs will have a counter. This example is identical to INTC: Hardware Vector Mode except for the differences between software and hardware vector modes. Note: eMIOS channel 0 interrupt vector numbers are different on MPC551x and MPC555x devices. Also, to generate the timed interrupt, the eMIOS channel mode will be the modulus counter for MPC555x devices that have that mode, or the newer modulus counter buffered mode. The ISRs will be written in C, so the appropriate registers will be saved and restored in the handler. The SPE will not be used, so its accumulator will not be saved in the stack frame of the interrupt handler. Exercise: Write a third interrupt handler that uses a software interrupt of a higher priority than the others. Invoke this new software interrupt from one of the other ISRs. MPC5500 Crystal 8 MHz Clocks and PLL eMIOS SIU_SYSCLK (MPC551x only) eMIOS Prescaler eMIOS Channel 0 Divide sysclk by 1 for eMIOS (Divide by 12) Mod. Counter: (Divide by 1 & count to 1000) Default sysclk: 16 MHz (MPC551x) 12 MHz (MPC555x) 8 MHz (MPC563x) Software settable interrupt req. 4 (set in eMIOS channel 0 interrupt service routine) Interrupt Request to INTC Interrupt Request to INTC INTC (Software Vector Mode) Common INTC interrupt request to CPU core Figure 8. Software Vector Mode Example Qorivva Simple Cookbook, Rev. 4 22 Freescale Semiconductor 4.2 Design The overall program flow for MPC551x is shown below. When an interrupt occurs, it is routed to either or both cores, as defined in the INTC_PSR for that vector. In this example only one processor is selected in the MPC551x for interrupts, processor 0 (e200z1). The selection is defined in INTC_PSR for each enabled interrupt, which here is eMIOS channel 0 and software interrupt 4. For MPC551x, there can be a different ISR Vector table for each core because each core has its own special purpose register, IVPR. The Vector Table Base Address is defined in each core’s INTC_ACKR, that is, either INTC_ACK_PRC0 for e200z1 and INTC_IACK_PRC1 for e200z0. This example uses only the e200z1 core. External input interrupt taken for selected processor vector to IVPR0:19 + 4 (for IVOR4) 16 bytes main.c file ivor_branch_table.s file IVOR Branch Table b IVOR0trap b IVOR1trap ... b IVOR4Handler ... IntcIsrVectors.c file ISR Vector Table: ... SW IRQ 4 ISR addr ... eMIOS Ch 0 ISR addr ... ... ... handlers.s file IVOR4Handler: prologue: save SRR0:1, fetch ISR vector, re-enable interrupts save registers branch to fetched ISR vector epilogue (return from ISR): restore most registers ensure prior stores completed disable interrupts write to EOIR to restore priority restore SRR0:1 return main { init IVPR, IVOR4 init INTC init eMIOS Chan 0 init SW 4 interrupt enable interrupts wait forever } emiosCh0ISR { increment counter if odd count, invoke SW 4 IRQ clear flag } SwIrq4ISR { increment counter clear flag } Figure 9. MPC551x Program Flow Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 23 The overall program flow for MPC555x is shown below. External input interrupt taken vector to IVPR0:15 + IVOR416:27 handlers.s file IntcIsrVectors.c file ISR Vector Table: ... SW IRQ 4 ISR addr ... eMIOS Ch 0 ISR addr ... ... ... IVOR4Handler: prologue: save SRR0:1, fetch ISR vector, re-enable interrupts save registers branch to fetched ISR vector epilogue (return from ISR): restore most registers ensure prior stores completed disable interrupts write to EOIR to restore priority restore SRR0:1 return main.c file main { init IVPR, IVOR4 init INTC init eMIOS Chan 0 init SW 4 interrupt enable interrupts wait forever } emiosCh0ISR { increment counter if odd count, invoke SW 4 IRQ clear flag } SwIrq4ISR { increment counter clear flag } Figure 10. MPC555x Program Flow Qorivva Simple Cookbook, Rev. 4 24 Freescale Semiconductor 4.2.1 Initialization Table 12. Initialization: INTC in Software Vector Mode Pseudo Code Step Relevant Bit Fields MPC551x Variable Counter for eMIOS channel 0 interrupts Initializations Counter for software 4 interrupts int emiosCh0Ctr = 0 int SWirq4Ctr = 0 Table Load INTC’s ISR vector table with addresses for: Initializations INTC Vector 4 ISR (SW interrupt request 4) INTC Vector 51 ISR (eMIOS channel 0) &SwIrq4ISR &emiosCh0ISR initIrqVectors Load the common interrupt vector prefix to spr IVPR spr IVPR =__IVPR_VALUE MPC555x: Initialize external input interrupt offset to the lower half of IVROR4 handler’s address initINTC initEMIOS – Initialize INTC: • Configure for software vector mode (HVEN=0) • Keep vector offset (VTES) at four bytes (MPC551x note: only proc’r 0, e200z1 is used here) HVEN_PRC0=0(551x) VTES_PRC0=0(551x) HVEN = 0 (MPC555x) VTES = 0 (MPC555x) Initialize ISR vector table base address per link file (MPC551x note: only proc’r 0, e200z1, is used here) VTBA = passed from linker Initialize eMIOS global settings for all channels: • Prescale sysclk by 12 for eMIOS clk • Enable eMIOS clock • Enable global time base • Enable stopping channels in debug mode GPRE = 11 (0xB) GPREN = 1 GTBE = 1 FRZ=1 MPC551x: Ensure channel is not disabled initSwIrq4 spr IVOR4 = IVOR4Handler @l INTC_MCR = 0 INTC_IACKR INTC_IACKR _PRC 0 = = __IACKR_ __IACKR_ VTBA_VALUE VTBA_VALUE EMIOS_MCR = 0x3400 B000 EMIOS_ UCDIS = 0 – Channel 0 Interrupt: raise priority to 1 • MPC551x: Select processor(s) to interrupt PRI = 1 INTC_PSR[58] INTC_PSR[51] PRC_SEL=0 (e200z1) = 0x01 = 0x01 Channel 0: set max count = 1,000 clks ( ~1 s each) A = 999 Channel 0: Enable channel as up counter & enable IRQ • Mode (MPC551x, 563x) = mod. up counter buffered • Mode (MPC555x) = modulus up counter • Counter bus select = internal counter • Channel prescaler = 1 • Enable channel prescaler • Enable freezing count in debug mode • Assign flag to assert IRQ (instead of DMA) • Enable flag to request IRQ MODE=0x50 or MODE=0x10 BSL=3 UCPRE=0 (default) UCPREN=1 FREN=1 DMA=0 (default) FEN=1 Software Interrupt 4: raise priority to 2 • MPC551x: Specify which processor(s) to interrupt PRI = 2 PRC_SEL=0 (e200z1) enableExtIrq Enable recognition of requests to INTC by lowering the PRI = 0 INTC’s current priority to 0 from default of 15 waitloop MPC555x EMIOS_A[0] = 999 EMIOS_ CCR[0] = 0x8202 0650 MPC555x: EMIOS_ CCR[0] = 0x8202 0610 MPC563x: EMIOS_ CCR[0] = 0x8202 0650 INTC_PSR[4] = 0x02 INTC_CPR _PRC0[PRI]= 0 INTC_CPR [PRI]= 0 Enable external interrupts by setting MSR[EE] = 1 wrteei 1 wait forever while 1 Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 25 4.2.2 Interrupt Handler Table 13. Stack Frame for INTC Interrupt Handler Stack Frame Area 32-bit GPRs Register Location in Stack r12 sp + 0x4C r11 sp + 0x48 r10 sp + 0x44 r9 sp + 0x40 r8 sp + 0x3C r7 sp + 0x38 r6 sp + 0x34 r5 sp + 0x30 r4 sp + 0x2C r3 sp + 0x28 r0 sp + 0x24 CR CR sp + 0x20 locals and padding XER sp + 0x1C CTR sp + 0x18 LR sp + 0x14 SRR1 sp + 0x10 SRR0 sp + 0x0C padding sp + 0x08 LR area LR placeholder sp + 0x04 back chain r1 sp Qorivva Simple Cookbook, Rev. 4 26 Freescale Semiconductor Table 14. IVOR4 Interrupt Handler (INTC in Software Vector Mode) (MPC555x: Must be 16 byte aligned and within 64KB of address in IVPR. MPC551x does not require alignment because it is the destination of a branch instruction.) Relevant Bit Fields Step prolog Pseudo Code MPC551x Create stack frame MPC555x stwu sp, –0x50 (sp) Save SRR0:1 because nested interrupts will be allowed store SRR0:1, r3 registers to stack frame Read pointer into ISR Vector Table Re-enable external interrupts by setting MSR[EE] r3 = INTC_IACKR_PRC0 EE = 1 wrteei 1 Read ISR address from ISR Vector Table and store into LR lwz r3, 0x0 (r3) mtLR r3 Save other appropriate registers for C ISR store other registers to stack frame Branch to ISR, saving return address epilog blrl Restore registers from stack frame except SRR0:1 and two working registers load most registers from stack frame Ensure interrupt flag in ISR has completed clearing before writing to INTC_EOIR Disable external interrupts by clearing MSR[EE] mbar EE = 0 Restore former INTC’s Current Priority wrteei 0 write 0 to INTC_EOIR_PRC0 Restore SRR0:1 and working registers write 0 to INTC_EOIR restore SRR0:1, working registers from stack frame Restore stack frame space addi sp, sp, 0x50 Return 4.2.3 r3 = INTC_IACKR rfi Interrupt Service Routines (ISRs) Table 15. ISR for eMIOS Channel 0 Step Relevant Bit Fields emiosCh0ISR Increment emiosCh0Ctr Pseudo Code emiosCh0Ctr ++ If emiosCh0Ctr is even, invoke Software Interrupt 4 SET = 1 Clear eMIOS Ch 0 interrupt flag by writing 1 to it FLAG = 1 if ((emosCh0Ctr&1)), INTC_SSCIR[4] = 2 EMIOS_CSR0[FLAG] = 1 Table 16. ISR for Software Interrupt 4 Step Relevant Bit Fields swIRQ4ISR Increment SWirq4ctr Clear Software IRQ 4 Pseudo Code SWirq4ctr++ CLR = 1 INTC_SSCIR[4] = 1 Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 27 4.3 4.3.1 /* /* /* /* /* /* /* /* /* /* /* /* /* /* /* Code main.c file main.c - Software vector mode program using C isr */ Jan 15, 2004 S.Mihalik -- Copyright Freescale, 2004. All Rights Reserved */ Feb 9, 2004 S.M. - removed unused cache init & enabled eMIOS FREEZE */ Mar 4, 2004 S.M. - added software interrupt 7 code */ May 19,2006 S.M.- renamed SWirq7Ctr to SWirq4Ctr for consistency */ Aug 12 2006 S.M. - made i volatile (to get fewer Nexus messages in loop) */ Jul 17 2007 SM - Passed IVPR value from link file, used relevant names */ for MPC551x bit fields & registers, invoked SW interrupt*/ on even count eMIOS Ch 0 ISRs, changed timeout to 1 msec*/ and changed EMIOS_MCR[PRE] & EMIOS Chan 0 A Register values */ Copyright Freescale Semiconductor, In.c 2007 All rights reserved. */ Notes: */ 1. MMU not initialized; must be done by debug scripts or BAM */ 2. SRAM not initialized; must be done by debug scripts or in a crt0 type file */ 3. Cache is not used */ #include "mpc563m.h" /* Use proper include file like mpc5510.h or mpc5554.h */ extern IVOR4Handler(); extern uint32_t __IVPR_VALUE; /* Interrupt Vector Prefix value from link file*/ extern const vuint32_t IntcIsrVectorTable[]; int emiosCh0Ctr = 0; /* Counter for eMIOS channel 0 interrupts */ int SWirq4Ctr = 0; /* Counter for software interrupt 4 */ asm void initIrqVectors(void) { lis r3, __IVPR_VALUE@h /* IVPR value is passed from link file */ ori r3, r3, __IVPR_VALUE@l mtivpr r3 /* The following two lines are required for MPC555x, and are not used for MPC551x*/ li r3, IVOR4Handler@l /* IVOR4 = lower half of handler address */ mtivor4r3 } void initINTC(void) { /* Use the first 3 or next 3 lines: */ Use the first 3 or next 3 lines for MPC551x or MPC555x: */ INTC.MCR.B.HVEN_PRC0 = 0;*/ /* MPC551x Proc'r 0: initialize for SW vector mode */ INTC.MCR.B.VTES_PRC0 = 0;*/ /* MPC551x Proc'r 0: default vector table 4B offsets*/ ITC.IACKR_PRC0.R= (uint32_t) &IntcIsrVectorTable[0];*//*MPC551x ISR table base*/ INTC.MCR.B.HVEN = 0; /* MPC555x: initialize for SW vector mode */ INTC.MCR.B.VTES = 0; /* MPC555x: Use default vector table 4B offsets */ INTC.IACKR.R = (uint32_t) &IntcIsrVectorTable[0]; /* MPC555x INTC ISR table base */ /* /* /* /* } void initEMIOS(void) { EMIOS.MCR.B.GPRE= 11; /* Divide sysclk by (11+1) for eMIOS clock */ EMIOS.MCR.B.GPREN = 1; /* Enable eMIOS clock */ EMIOS.MCR.B.GTBE = 1; /* Enable global time base */ EMIOS.MCR.B.FRZ = 1; /* Enable stopping channels when in debug mode */ /* Following for MPC551x only: */ /* EMIOS.UCDIS.R = 0; */ /* Ensure all channels are enabled */ /* Use one of the following two lines: */ /*INTC.PSR[58].R = 1; */ /* MPC551x: Raise eMIOS chan 0 IRQ priority = 1 */ INTC.PSR[51].R = 1; /* MPC555x: Raise eMIOS chan 0 IRQ priority = 1 */ EMIOS.CH[0].CADR.R = 999; /* Period will be 999+1 = 1000 channel clocks */ /* Use one of the next two lines: MCB mode is not in all MPC555x devices */ EMIOS.CH[0].CCR.B.MODE = 0x50; /* MPC551x or MPC563x: Mod. Ctr. Buffered (MCB) */ /*EMIOS.CH[0].CCR.B.MODE = 0x10;*/ /* MPC555x: Modulus Counter (MC), internal clk */ EMIOS.CH[0].CCR.B.BSL = 0x3; /* Use internal counter */ EMIOS.CH[0].CCR.B.UCPREN = 1; /* Enable prescaler; uses default divide by 1 */ EMIOS.CH[0].CCR.B.FREN = 1; /* Freeze channel registers when in debug mode */ EMIOS.CH[0].CCR.B.FEN=1; /* Flag enables interrupt */ } Qorivva Simple Cookbook, Rev. 4 28 Freescale Semiconductor void initSwIrq4(void) { INTC.PSR[4].R = 2; } /* Software interrupt 4 IRQ priority = 2 */ void enableIrq(void) { /* Use one of the following two lines to lower the INTC current priority */ /*INTC.CPR_PRC0.B.PRI = 0; */ /* MPC551x Proc'r 0: Lower INTC's current priority */ INTC.CPR.B.PRI = 0; /* MPC555x: Lower INTC's current priority */ asm(" wrteei 1"); /* Enable external interrupts */ } void main (void) { vuint32_t i = 0; initIrqVectors(); initINTC(); initEMIOS(); initSwIrq4(); enableIrq(); } /* Dummy idle counter */ /* /* /* /* /* Initialize exceptions: only need to load IVPR */ Initialize INTC for software vector mode */ Initialize eMIOS channel 0 for 1kHz IRQ, priority 2 */ Initialize software interrupt 4 */ Ensure INTC current prority=0 & enable IRQ */ while (1) { i++; } void emiosCh0ISR(void) { emiosCh0Ctr++; if ((emiosCh0Ctr & 1)==0) { INTC.SSCIR[4].R = 2; } EMIOS.CH[0].CSR.B.FLAG=1; } void SwIrq4ISR(void) { SWirq4Ctr++; INTC.SSCIR[4].R = 1; } /* Increment interrupt counter */ /* If emiosCh0Ctr is even*/ /* then nvoke software interrupt 4 */ /* Clear channel's flag */ /* Increment interrupt counter */ /* Clear channel's flag */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 29 4.3.2 # # # # # # # # handlers.s file handlers.s - INTC software vector mode example Description: Creates prolog, epilog for C ISR and enables nested interrupts Rev 1.0: April 23, 2004, S Mihalik, Rev 1.1 Aug 2, 2004 SM - delayed writing to EOIR until after disabling EE in epilog Rev 1.2 Sept 8 2004 SM - optimized & corrected r3,r4 restore sequence from rev 1.1 Rev 1.2 Sept 21 2004 SM - optimized by minimizing time interrupts are disabled Rev 1.3 Jul 2 2007 SM - Changes for MPC551x and mapped to .ivor_handlers section Copyright Freescale Semiconductor, Inc. 2007. All rights reserved # STACK FRAME DESIGN: Depth: 20 words (0xA0, or 80 bytes) # ************* ______________ # 0x4C * GPR12 * ^ # 0x48 * GPR11 * | # 0x44 * GPR10 * | # 0x40 * GPR9 * | # 0x3C * GPR8 * | # 0x38 * GPR7 * GPRs (32 bit) # 0x34 * GPR6 * | # 0x30 * GPR5 * | # 0x2C * GPR4 * | # 0x28 * GPR3 * | # 0x24 * GPR0 * ___v__________ # 0x20 * CR * __CR__________ # 0x1C * XER * ^ # 0x18 * CTR * | # 0x14 * LR * locals & padding for 16 B alignment # 0x10 * SRR1 * | # 0x0C * SRR0 * | # 0x08 * padding * ___v__________ # 0x04 * resvd- LR * Reserved for calling function # 0x00 * SP * Backchain (same as gpr1 in GPRs) # ************* .section .ivor_handlers .globl IVOR4Handler .align 16 # Align IVOR handlers on a 16 byte boundary for MPC555x # GHS, Cygnus, Diab(default) use .align 4; CodeWarrior .align 16 .equ .equ .equ .equ INTC_IACKR_PRC0, 0xfff48010# INTC_EOIR_PRC0, 0xfff48018 # INTC_IACKR, 0xfff48010 # INTC_EOIR, 0xfff48018 # IVOR4Handler: prolog: stwu r1, -0x50 (r1) stw r3, 0x28 (r1) mfsrr0 r3 stw r3, 0x0C (r1) mfsrr1 r3 stw r3, 0x10 (r1) # # # # # # # # MPC551x: MPC551x: MPC555x: MPC555x: Proc'r 0 Interrupt Acknowledge Reg addr Proc'r 0 End Of Interrupt Reg. addr Interrupt Acknowledge Reg. addr. End Of Interrupt Reg. addr. PROLOGUE Create stack frame and store back chain Store a working register Store SRR0:1 (must be done before enabling EE) The following two lines are for MPC551x processors: */ lis r3, INTC_IACKR_PRC0@ha # Read proc'0 ptr into ISR Vector Table, store in r3 lwz r3, INTC_IACKR_PRC0@l(r3) The following two lines are for MPC555x processors: */ lis r3, INTC_IACKR@ha # Read pointer into ISR Vector Table & store in r3 lwz r3, INTC_IACKR@l(r3) lwz r3, 0x0(r3) # Read ISR address from ISR Vector Table using pointer wrteei stw mflr stw mtlr 1 r4, r4 r4, r3 # Set MSR[EE]=1(must wait a couple clocks after reading IACKR) 0x2C (r1) # Store a second working register # Store LR (LR will be used for ISR Vector) 0x14 (r1) # Store ISR address to LR to use for branching later Qorivva Simple Cookbook, Rev. 4 30 Freescale Semiconductor stw stw stw stw stw stw stw stw stw mfcr stw mfxer stw mfctr stw r12, r11, r10, r9, r8, r7, r6, r5, r0, r3 r3, r3 r3, r3 r3, 0x4C 0x48 0x44 0x40 0x3C 0x38 0x34 0x30 0x24 (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) 0x20 (r1) 0x1C (r1) 0x18 (r1) blrl # Store rest of gprs # Store CR # Store XER # Store CTR # Branch to ISR, but return here epilog: # EPILOGUE lwz r3, 0x14 (r1) # Restore LR mtlr r3 lwz r3, 0x18 (r1) # Restore CTR mtctr r3 lwz r3, 0x1C (r1) # Restore XER mtxer r3 lwz r3, 0x20 (r1) # Restore CR mtcrf 0xff, r3 lwz r0, 0x24 (r1) # Restore other gprs except working registers lwz r5, 0x30 (r1) lwz r6, 0x34 (r1) lwz r7, 0x38 (r1) lwz r8, 0x3C (r1) lwz r9, 0x40 (r1) lwz r10, 0x44 (r1) lwz r11, 0x48 (r1) lwz r12, 0x4C (r1) mbar 0 # Ensure store to clear interrupt's flag bit completed # The following line is for the MPC551x: # lis r3, INTC_EOIR_PRC0@ha# MPC551x: Load upper half of proc'r 0 EIOR addr to r3 # The following line is for the MPC555x: lis r3, INTC_EOIR@ha# MPC555x: Load upper half of EIOR address to r3 li r4, 0 wrteei 0 # Disable interrupts for rest of handler # The following line is for MPC551x: # stw r4, INTC_EOIR_PRC0@l(r3)# MPC551x: Write 0 to proc'r 0 INTC_EOIR # The following line is for MPC555x: stw r4, INTC_EOIR@l(r3)# MPC555x: Write 0 to INTC_EOIR lwz r3, 0x0C (r1) # Restore SRR0 mtsrr0 r3 lwz r3, 0x10 (r1) # Restore SRR1 mtsrr1 r3 lwz r4, 0x2C (r1) # Restore working registers lwz r3, 0x28 (r1) addi r1, r1, 0x50 # Delete stack frame rfi # End of Interrupt Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 31 4.3.3 /* /* /* /* /* /* /* /* /* /* IntcIsrVectors.c file IntcIsrVectors.c - table of ISRs for INTC in SW vector Mode */ Description: Contains addresses for 310 ISR vectors */ Table address gets loaded to INTC_IACKR */ Alignment: MPC551x: 2 KB after a 4KB boundary; MPC555x: 64 KB*/ Copyright Freescale Semiconductor Inc 2007. All rights reserved. */ April 22, 2004 S. Mihalik */ March 16, 2006 S. Mihalik - Modified for compile with Diab 5.3 */ Jun 29 2006 SM - Used pragma align instead of hard coding address */ Jul 5 2007 SM - alignment now done in link file; changes for MPC551x */ Aug 30 2007 SM - Added pragma for CodeWarrior */ #include "mpc563m.h" /* Use proper include file like mpc5510.h or mpc5554.h */ void dummy (void); extern void SwIrq4ISR(void); extern void emiosCh0ISR(void); /* Use pragma next two lines with CodeWarrior compile */ #pragma section data_type ".intc_sw_isr_vector_table" ".intc_sw_isr_vector_table" data_mode=far_abs uint32_t IntcIsrVectorTable[] = { /* Use next two lines with Diab compile */ /* #pragma section CONST ".intc_sw_isr_vector_table" */ /* Diab compiler pragma */ /* const uint32_t IntcIsrVectorTable[] = { */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&SwIrq4ISR, /* ISRs 00 - 04 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 05 - 09 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 10 - 14 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 15 - 19 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 20 - 24 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 25 - 29 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 30 - 34 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 35 - 39 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 40 - 44 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 45 - 49 */ /* Use next 2 lines for MPC551x: */ /*(uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, */ /* ISRs 50 - 54 */ /*(uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&emiosCh0ISR, (uint32_t)&dummy, */ /* ISRs 55 - 59 */ /* Use next 2 lines for MPC555x: */ (uint32_t)&dummy, (uint32_t)&emiosCh0ISR, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 50 - 54 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 55 - 59 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 60 - 64 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 55 - 69 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 70 - 74 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 75 - 79 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 80 - 84 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 85 - 89 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 90 - 94 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 95 - 99 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* /* /* /* /* /* /* /* /* /* /* /* /* /* /* /* /* /* /* /* ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs 100 105 110 115 120 125 130 135 140 145 150 155 160 155 170 175 180 185 190 195 - 104 109 114 119 124 129 134 139 144 149 154 159 164 169 174 179 184 189 194 199 */ */ */ */ */ */ */ */ */ */ */ */ */ */ */ */ */ */ */ */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* /* /* /* /* /* /* /* ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs 200 205 210 215 220 225 230 235 - 204 209 214 219 224 229 234 239 */ */ */ */ */ */ */ */ Qorivva Simple Cookbook, Rev. 4 32 Freescale Semiconductor (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* /* /* /* /* /* /* /* /* /* /* /* ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs ISRs 240 245 250 255 260 255 270 275 280 285 290 295 - 244 249 254 259 264 269 274 279 284 289 294 299 */ */ */ */ */ */ */ */ */ */ */ */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 300 - 304 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 305 - 309 */ }; void dummy (void) { while (1) {}; /* Wait forever or for watchdog timeout */ } Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 33 4.3.4 # # # # # ivor_branch_table.s file (MPC551x only) ivor_branch_table.s - for use with MPC551x only Description: Branch table for 16 MPC551x core interrupts Copyright Freescale 2007. All Rights Reserved Rev 1.0 Jul 6 2007 S Mihalik - Initial version Rev 1.1 Aug 30 2007 SM - Made IVOR4Handler extern .extern IVOR4Handler .section .ivor_branch_table .equ SIXTEEN_BYTES, 16 IVOR0trap: IVOR1trap: IVOR2trap: IVOR3trap: b IVOR5trap: IVOR6trap: IVOR7trap: IVOR8trap: IVOR9trap: IVOR10trap: IVOR11trap: IVOR12trap: IVOR13trap: IVOR14trap: IVOR15trap: # 16 byte alignment required for table entries # Diab compiler uses value of 4 (2**4=16) # CodeWarrior, GHS, Cygnus use 16 .align SIXTEEN_BYTES b IVOR0trap # IVOR 0 interrupt handler .align SIXTEEN_BYTES b IVOR1trap # IVOR 1 interrupt handler .align SIXTEEN_BYTES b IVOR2trap # IVOR 2 interrupt handler .align SIXTEEN_BYTES b IVOR3trap # IVOR 3 interrupt handler .align SIXTEEN_BYTES IVOR4Handler # IVOR 4 interrupt handler (External Interrupt) .align SIXTEEN_BYTES b IVOR5trap # IVOR 5 interrupt handler .align SIXTEEN_BYTES b IVOR6trap # IVOR 6 interrupt handler .align SIXTEEN_BYTES b IVOR7trap # IVOR 7 interrupt handler .align SIXTEEN_BYTES b IVOR8trap # IVOR 8 interrupt handler .align SIXTEEN_BYTES b IVOR9trap # IVOR 9 interrupt handler .align SIXTEEN_BYTES b IVOR10trap # IVOR 10 interrupt handler .align SIXTEEN_BYTES b IVOR11trap # IVOR 11 interrupt handler .align SIXTEEN_BYTES b IVOR12trap # IVOR 12 interrupt handler .align SIXTEEN_BYTES b IVOR13trap # IVOR 13 interrupt handler .align SIXTEEN_BYTES b IVOR14trap # IVOR 14 interrupt handler .align SIXTEEN_BYTES b IVOR15trap # IVOR15 interrupt handler Qorivva Simple Cookbook, Rev. 4 34 Freescale Semiconductor 5 INTC: Hardware Vector Mode 5.1 Description Task: Using the Interrupt Controller (INTC) in hardware vector mode, this program provides two interrupts that show nesting. The first interrupt is generated from an eMIOS channel at a 1 kHz rate using the MPC555x default system clock. For the MPC551x, the faster default system clock produces an eMIOS channel timeout rate of approximately 1 kHz 16/12 = 1.33 kHz. The interrupt handler will re-enable interrupts and later, in its interrupt service routine (ISR), invoke a second interrupt every other time,. This provides an approximate 1 millisecond task (ISR) from the eMIOS channel and an approximate 2 millisecond task (ISR) from the software interrupt. The software interrupt will have a higher priority, so the one eMIOS ISR is preempted by the software interrupt. Both ISRs will have a counter. This example is identical to INTC: Software Vector Mode, except for the differences between software and hardware vector modes. Note: eMIOS channel 0 interrupt vector numbers are different on MPC551x and MPC555x devices. Also, to generate the timed interrupt, the eMIOS channel mode will be the modulus counter for MPC555x devices that have that mode, or the newer modulus counter buffered mode. The ISRs will be written in C, so the appropriate registers will be saved and restored in the handler. The SPE will not be used, so its accumulator will not be saved in the stack frame of the interrupt handler.1 Exercise: Write a third interrupt handler that uses a software interrupt of a higher priority than the others. Invoke this new software interrupt from one of the other ISRs. MPC5500 Crystal 8 MHz Clocks and PLL eMIOS SIU_SYSCLK (MPC551x only) eMIOS Prescaler Divide sysclk by 1 for eMIOS (Divide by 12) Default sysclk: 16 MHz (MPC551x) 12 MHz (MPC555x) 8 MHz (MPC563x) eMIOS Channel 0 Mod. Counter: (Divide by 1 & count to 1000) Software settable interrupt req. 4 (set in eMIOS channel 0 interrupt service routine) Interrupt Request to INTC Interrupt Request to INTC INTC (Hardware Vector Mode) Unique INTC interrupt request to CPU core Figure 11. Hardware Vector Mode Example 1. The SPE’s accumulator is not saved in the prologue. It is assumed SPE instructions are not used in this example. Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 35 5.2 Design The overall program flow is shown below, followed by design steps and a stack frame implementation. The INTC when used in hardware vector mode uses a branch table for getting to each INTC vector’s handler. These are shown as handler_0, handler_1, etc. (Note: the code in this example does not have handlers for each INTC vector, so a common dummy trap address is used.) For MPC551x, the second core has its own special purpose register, IVPR, so the second core would have its own IntcHanderBranchTable (not included in this example). Also for MPC551x, either or both processors can receive the interrupt request from the Interrupt Controller. In this example only one processor is selected in the MPC551x, processor 0 (e200z1). The selection is defined in INTC_PSR for each enabled interrupt, which here is eMIOS channel 0 and software interrupt 4. INTC vector for eMIOS Channel 0 interrupt taken vector to MPC551x: IVPR0:19 + 0x800 + 51 (0x4) MPC555x: IVPR0:15 + 51 (0x10) handlers.s file intc_hw_branch_table.s file IntcHandlerBranchTable b handler_0 ... b handler_1 ... b handler_2 ... b handler_3 ... b SwIrq4Handler ... ... ... ... b emiosCh0Handler ... ... ... main.c file emiosCh0Handler: prologue: save registers re-enable interrupts branch to emiosCh0ISR epilogue (return from ISR): restore most registers ensure prior stores completed disable interrupts write to EOIR to restore priority restore SRR0:1 return SwIrq4Handler: prologue: save registers re-enable interrupts branch to SwIrq4ISR epilogue (return from ISR): restore most registers ensure prior stores completed disable interrupts write to EOIR to restore priority restore SRR0:1 return main { init IVPR init INTC init SwIrq4 init eMIOS Chan 0 enable interrupts wait forever } emiosCh0ISR { increment counter if odd count, invoke SW 4 IRQ clear flag } SwIrq4ISR { increment counter clear flag } Figure 12. Overall Program Flow (Shown for MPC555x which uses INTC vector 51 for eMIOS channel 0. MPC551x uses INTC vector 58 for eMIOS channel 0.) Qorivva Simple Cookbook, Rev. 4 36 Freescale Semiconductor 5.2.1 Initialization Table 17. Initialization: INTC in Hardware Vector Mode Pseudo Code Step Relevant Bit Fields MPC551x Variable Counter for eMIOS channel 0 interrupts Initializations Counter for software 4 interrupts MPC555x int emiosCh0Ctr = 0 int SWirq4Ctr = 0 Table Load INTC HW Branch Table with ISR names for: Initializations INTC Vector 4 ISR (SW interrupt request 4) INTC Vector 51 ISR (eMIOS channel 0) b SwIrq4ISR b emiosCh0ISR initIrqVectors Load the common interrupt vector prefix to spr IVPR. Value is defined in the link file. spr IVPR =__IVPR_VALUE initINTC Initialize INTC: • Configure for hardware vector mode (MPC551x note: only proc’r 0, e200z1, is used here) HVEN_PRC0 = 0(551x) INTC_MCR INTC_MCR HVEN = 0 (MPC555x) [HVEN_PRC0] [HVEN] =1 =1 initEMIOS Initialize EMIOS global settings for all channels: • Prescale sysclk by 12 for eMIOS clock • Enable eMIOS clock • Enable global time base • Enable stopping channels in debug mode GPRE = 11 (0xB) GPREN = 1 GTBE = 1 FRZ = 1 MPC551x: Ensure channel is not disabled initSwIrq4 Channel 0 Interrupt: • Raise priority to 1 • MPC551x: Select processor(s) to interrupt PRI = 1 PRC_SEL = 0 (e200z1) Channel 0: set max count = 1,000 clks ( ~ 1 s each) A = 999 Channel 0: Enable channel as up counter & enable IRQ • Mode (MPC551x, 563x) = modulus up counter buffered • Mode (MPC555x) = modulus up counter • Counter bus select = internal counter • Channel prescaler = 1 • Enable channel prescaler • Enable freezing count in debug mode • Assign flag to assert IRQ (instead of DMA) • Enable flag to request IRQ MODE = 0x50 or MODE = 0x10 BSL = 3 UCPRE = 0 (default) UCPREN = 1 FREN = 1 DMA = 0 (default) FEN = 1 Software Interrupt 4: • Raise priority to 2 • MPC551x: Specify which processor(s) to interrupt EMIOS_ UCDIS = 0 – INTC_PSR [58] = 0x01 INTC_PSR [51] = 0x01 EMIOS_A[0] = 999 EMIOS_ CCR[0] = 0x8202 0650 MPC555x: EMIOS_ CCR[0] = 0x8202 0610 MPC563x: EMIOS_ CCR[0] = 0x8202 0650 INTC_PSR[4] = 0x02 PRI = 2 PRC_SEL = 0 (e200z1) enableExtIrq Enable recognition of requests to INTC by lowering the PRI = 0 INTC’s current priority to 0 from default of 15 (MPC551x note: only proc’r 0, e200z1, is used here) waitloop EMIOS_MCR = 0x3400 B000 INTC_CPR _PRC0[PRI] =0 INTC_CPR [PRI] =0 Enable external interrupts by setting MSR[EE] = 1 wrteei 1 wait forever while 1 Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 37 5.2.2 Interrupt Handlers Table 18. Stack Frame for INTC Interrupt Handler (Same as for Software Vector Mode Example) Stack Frame Area 32-bit GPRs Register Location in Stack r12 sp + 0x4C r11 sp + 0x48 r10 sp + 0x44 r9 sp + 0x40 r8 sp + 0x3C r7 sp + 0x38 r6 sp + 0x34 r5 sp + 0x30 r4 sp + 0x2C r3 sp + 0x28 r0 sp + 0x24 CR CR sp + 0x20 locals and padding XER sp + 0x1C CTR sp + 0x18 LR sp + 0x14 SRR1 sp + 0x10 SRR0 sp + 0x0C padding sp + 0x08 LR area LR placeholder sp + 0x04 back chain r1 sp Qorivva Simple Cookbook, Rev. 4 38 Freescale Semiconductor Table 19. eMIOS Channel 0 Interrupt Handler (INTC in Hardware Vector Mode) Relevant Bit Fields Step prolog Pseudo Code MPC551x Create stack frame stwu sp, – 0x50 (sp) Save SRR0:1 because nested interrupts will be allowed Re-enable external interrupts by setting MSR[EE] store SRR0:1, r3 registers to stack frame EE = 1 wrteei 1 Save other appropriate registers for C ISR store other registers to stack frame Branch to ISR, saving return address emiosCh0 Increment emiosCh0Ctr ISR If emiosCh0Ctr is even, invoke software Interrupt 4 bl emiosCh0ISR emiosCh0Ctr++ if ((emiosCh0Ctr&1), INTC_SSCIR[4] = 1 SET = 1 Clear eMIOS Ch 0 interrupt flag by writing 1 to it epilog MPC555x FLAG = 1 EMIOS_CSR0[FLAG] = 1 Restore registers from stack frame except SRR0:1 and two working registers load most registers from stack frame Ensure interrupt flag in ISR has completed clearing before writing to INTC_EOIR Disable external interrupts by clearing MSR[EE] Restore former INTC’s current priority Restore SRR0:1 and working registers mbar EE = 0 wrteei 0 write 0 to INTC_EOIR_PRC0 write 0 to INTC_EOIR restore SRR0:1, working registers from stack frame Restore stack frame space addi sp, sp, 0x50 Return rfi Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 39 Table 20. Software 4 Interrupt Handler (INTC in Hardware Vector Mode) Relevant Bit Fields Step prolog Pseudo Code MPC551x Create stack frame MPC555x stwu sp, – 0x50 (sp) Save SRR0:1 because nested interrupts will be allowed Re-enable external interrupts by setting MSR[EE] store SRR0:1, r3 registers to stack frame EE = 1 wrteei 1 Save other appropriate registers for C ISR store other registers to stack frame Branch to ISR, saving return address bl SWIrq4ISR SWIrq4 ISR Increment SWirq4Ctr SWirq4Ctr++ epilog Restore registers from stack frame except SRR0:1 and two working registers Clear software IRQ 4 interrupt flag by writing 1 to it CLR = 1 INTC_SSCIR4 = 1 load most registers from stack frame Ensure interrupt flag in ISR has completed clearing before writing to INTC_EOIR Disable external interrupts by clearing MSR[EE] Restore former INTC’s current priority Restore SRR0:1 and working registers mbar EE = 0 wrteei 0 write 0 to INTC_EOIR_PRC0 write 0 to INTC_EOIR restore SRR0:1, working registers from stack frame Restore stack frame space addi sp, sp, 0x50 Return rfi Qorivva Simple Cookbook, Rev. 4 40 Freescale Semiconductor 5.3 5.3.1 /* /* /* /* /* /* /* /* /* /* /* /* /* Code main.c file main.c - INTC in Hardware vector mode using C ISRs */ Copyright Freescale Semiconductor, Inc. 2007. All Rights Reserved */ Rev 0.1 Oct 1 2004 Steve Mihalik */ Rev 1.0 May 19,2006 S.M.- renamed SWirq7Ctr to SWirq4Ctr for consistency */ Rev 1.1 Aug 12 1006 SM - Made i volatile (to get fewer Nexus msgs in loop)*/ Jul 17 2007 SM - Passed IVPR value from link file, used relevant names */ for MPC551x bit fields & registers, invoked SW interrupt*/ on even count eMIOS Ch 0 ISRs, changed timeout to 1 msec*/ and changed EMIOS_MCR[PRE] & EMIOS Chan 0 A Register values */ Notes: */ 1. MMU not initialized; must be done by debug scripts or BAM */ 2. SRAM not initialized; must be done by debug scripts or in a crt0 type file */ 3. Cache is not used */ #include "mpc563m.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ extern uint32_t __IVPR_VALUE; /* Interrupt Vector Prefix value from link file*/ int emiosCh0Ctr = 0; /* Counter for eMIOS channel 0 interrupts */ int SWirq4Ctr = 0; /* Counter for software interrupt 4 */ asm void initIrqVectors(void) { lis r3, __IVPR_VALUE@h ori r3, r3, __IVPR_VALUE@l mtivpr r3 } /* IVPR value is passed from link file */ /* Note: IVPR lower bits are unused in MPC555x */ void initINTC(void) { /* Use one of the next two lines: */ /*INTC.MCR.B.HVEN_PRC0 = 1; */ /* MPC551x: Proc'r 0: initialize for HW vector mode*/ INTC.MCR.B.HVEN = 1; /* MPC555x: initialize for HW vector mode */ } void initEMIOS(void) { EMIOS.MCR.B.GPRE= 11; /* Divide sysclk by (11+1) for eMIOS clock */ EMIOS.MCR.B.GPREN = 1; /* Enable eMIOS clock */ EMIOS.MCR.B.GTBE = 1; /* Enable global time base */ EMIOS.MCR.B.FRZ = 1; /* Enable stopping channels when in debug mode */ /* Following for MPC551x only: */ /*EMIOS.UCDIS.R = 0; */ /* MPC551x: Ensure all channels are enabled */ /* Use one of the following two lines: */ /* INTC.PSR[58].R = 1; /* MPC551x: Raise eMIOS chan 0 IRQ priority = 1 */ INTC.PSR[51].R = 1; */ /* MPC555x: Raise eMIOS chan 0 IRQ priority = 1 */ EMIOS.CH[0].CADR.R = 999; /* Period will be 999+1=1000 channel clocks */ /* Use one of the following two lines for mode:*/ /*(Note: MCB mode is not in all MPC555x devices)*/ EMIOS.CH[0].CCR.B.MODE = 0x50; /* MPC551x or MPC563x: Mod. Counter Buffered (MCB*/ /*EMIOS.CH[0].CCR.B.MODE = 0x10;*/ /* MPC555x: Modulus Counter (MC) */ EMIOS.CH[0].CCR.B.BSL = 0x3; /* Use internal counter */ EMIOS.CH[0].CCR.B.UCPREN = 1; /*Enable prescaler; uses default divide by 1*/ EMIOS.CH[0].CCR.B.FREN = 1; /*Freeze channel registers when in debug mode*/ EMIOS.CH[0].CCR.B.FEN=1; /* Flag enables interrupt */ } void initSwIrq4(void) { INTC.PSR[4].R = 2;/* Software interrupt 4 IRQ priority = 2 */ } void enableIrq(void) { /* Use one of the following two lines: */ /*INTC.CPR_PRC0.B.PRI = 0; /* /* MPC551x Proc'r 0: Lower INTC's current priority*/ INTC.CPR.B.PRI = 0; /* MPC555x: Lower INTC's current priority */ asm(" wrteei 1"); /* Enable external interrupts */ } Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 41 void main (void) { vuint32_t i = 0; initIrqVectors(); initINTC(); initEMIOS(); initSwIrq4(); enableIrq(); } /* Dummy idle counter */ /* /* /* /* /* Initialize exceptions: only need to load IVPR */ Initialize INTC for hardware vector mode */ Initialize eMIOS channel 0 for 1kHz IRQ, priority 2 */ Initialize software interrupt 4 */ Ensure INTC current prority=0 & enable IRQ */ while (1) { i++; } void emiosCh0ISR(void) { emiosCh0Ctr++; /* Increment interrupt counter */ if ((emiosCh0Ctr & 1)==0) { /* If emiosCh0Ctr is even*/ INTC.SSCIR[4].R = 2; /* then nvoke software interrupt 4 */ } EMIOS.CH[0].CSR.B.FLAG=1; /* Clear channel's flag */ } void SwIrq4ISR(void) { SWirq4Ctr++; INTC.SSCIR[4].R = 1; } /* Increment interrupt counter */ /* Clear channel's flag */ Qorivva Simple Cookbook, Rev. 4 42 Freescale Semiconductor 5.3.2 # # # # # # # # handlers.s file handlers.s - INTC hardware vector mode example Description: Creates prolog, epilog for C ISR and enables nested interrupts Rev 1.0 Jan 5, 2004 S Mihalik Rev 1.1 Aug 2, 2004 SM - delayed writing to EOIR until after EE is disabled Rev 1.2 Jul 18, 2005 SM - .org, .section & .extern changes for CodeWarrior 1.5 Rev 1.3 Jul 6, 2007 SM - Moved isr_hw_bracnh_table to new file, added new 551x EOIR Rev 1.4 Aug 30 2007 SM - Added .text directive Copyright Freescale Semiconductor, Inc. 2007. All rights reserved # STACK FRAME DESIGN: Depth: 20 words (0xA0, or 80 bytes) # ************* ______________ # 0x4C * GPR12 * ^ # 0x48 * GPR11 * | # 0x44 * GPR10 * | # 0x40 * GPR9 * | # 0x3C * GPR8 * | # 0x38 * GPR7 * GPRs (32 bit) # 0x34 * GPR6 * | # 0x30 * GPR5 * | # 0x2C * GPR4 * | # 0x28 * GPR3 * | # 0x24 * GPR0 * ___v__________ # 0x20 * CR * __CR__________ # 0x1C * XER * ^ # 0x18 * CTR * | # 0x14 * LR * locals & padding for 16 B alignment # 0x10 * SRR1 * | # 0x0C * SRR0 * | # 0x08 * padding * ___v__________ # 0x04 * resvd- LR * Reserved for calling function # 0x00 * SP * Backchain (same as gpr1 in GPRs) # ************* .equINTC_EOIR,0xfff48018 # MPC555x: End Of Interrupt Reg. addr. .equINTC_EOIR_PRC0, 0xfff48018 # MPC551x: Proc'r 0 End Of Interrupt Reg. addr. .extern emiosCh0ISR .extern SwIrq4ISR .globl emiosCh0Handler .globl SwIrq4Handler .text Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 43 emiosCh0Handler: stwu r1, -0x50 (r1) stw mfsrr0 stw mfsrr1 stw r3, r3 r3, r3 r3, wrteei 1 stw stw stw stw stw stw stw stw stw stw mfcr stw mfxer stw mfctr stw mflr stw r12, r11, r10, r9, r8, r7, r6, r5, r4, r0, r3 r3, r3 r3, r3 r3, r3 r3, bl emiosCh0ISR 0x28 (r1) 0x0C (r1) # PROLOGUE # Create stack frame and store back chain # Store a working register # Store SRR0:1 (must be done before enabling EE) 0x10 (r1) # Set MSR[EE]=1 0x4C 0x48 0x44 0x40 0x3C 0x38 0x34 0x30 0x2C 0x24 (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) 0x20 (r1) 0x1C (r1) 0x18 (r1) 0x14 (r1) # Store rest of gprs # Store CR # Store XER # Store CTR # Store LR # EPILOGUE lwz r3, 0x14 (r1) # Restore LR mtlr r3 lwz r3, 0x18 (r1) # Restore CTR mtctr r3 lwz r3, 0x1C (r1) # Restore XER mtxer r3 lwz r3, 0x20 (r1) # Restore CR mtcrf 0xff, r3 lwz r0, 0x24 (r1) # Restore other gprs except working registers lwz r5, 0x30 (r1) lwz r6, 0x34 (r1) lwz r7, 0x38 (r1) lwz r8, 0x3C (r1) lwz r9, 0x40 (r1) lwz r10, 0x44 (r1) lwz r11, 0x48 (r1) lwz r12, 0x4C (r1) mbar 0 # Ensure store to clear flag bit has completed # The following line is for the MPC551x: # lis r3, INTC_EOIR_PRC0@ha# MPC551x: Load upper half of proc'r 0 EIOR addr to r3 # The following line is for the MPC555x: lis r3, INTC_EOIR@ha# MPC555x: Load upper half of EIOR address to r3 li r4, 0 wrteei 0 # Disable interrupts for rest of handler # The following line is for MPC551x: # stw r4, INTC_EOIR_PRC0@l(r3)# MPC551x: Write 0 to proc'r 0 INTC_EOIR # The following line is for MPC555x: stw r4, INTC_EOIR@l(r3)# MPC555x: Write 0 to INTC_EOIR lwz r3, 0x0C (r1) # Restore SRR0 mtsrr0 r3 lwz r3, 0x10 (r1) # Restore SRR1 mtsrr1 r3 lwz r4, 0x2C (r1) # Restore working registers lwz r3, 0x28 (r1) addi r1, r1, 0x50 # Delete stack frame rfi # End of Interrupt Qorivva Simple Cookbook, Rev. 4 44 Freescale Semiconductor SwIrq4Handler: stwu r1, -0x50 (r1) stw mfsrr0 stw mfsrr1 stw r3, r3 r3, r3 r3, wrteei 1 stw stw stw stw stw stw stw stw stw stw mfcr stw mfxer stw mfctr stw mflr stw r12, r11, r10, r9, r8, r7, r6, r5, r4, r0, r3 r3, r3 r3, r3 r3, r3 r3, bl SwIrq4ISR 0x28 (r1) 0x0C (r1) # PROLOGUE # Create stack frame and store back chain # Store a working register # Store SRR0:1 (must be done before enabling EE) 0x10 (r1) # Set MSR[EE]=1 0x4C 0x48 0x44 0x40 0x3C 0x38 0x34 0x30 0x2C 0x24 (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) 0x20 (r1) 0x1C (r1) 0x18 (r1) 0x14 (r1) # Store rest of gprs # Store CR # Store XER # Store CTR # Store LR # EPILOGUE lwz r3, 0x14 (r1) # Restore LR mtlr r3 lwz r3, 0x18 (r1) # Restore CTR mtctr r3 lwz r3, 0x1C (r1) # Restore XER mtxer r3 lwz r3, 0x20 (r1) # Restore CR mtcrf 0xff, r3 lwz r0, 0x24 (r1) # Restore other gprs except working registers lwz r5, 0x30 (r1) lwz r6, 0x34 (r1) lwz r7, 0x38 (r1) lwz r8, 0x3C (r1) lwz r9, 0x40 (r1) lwz r10, 0x44 (r1) lwz r11, 0x48 (r1) lwz r12, 0x4C (r1) mbar 0 # Ensure store to clear flag bit has completed # The following line is for the MPC551x: # lis r3, INTC_EOIR_PRC0@ha# MPC551x: Load upper half of proc'r 0 EIOR addr to r3 # The following line is for the MPC555x: lis r3, INTC_EOIR@ha# MPC555x: Load upper half of EIOR address to r3 li r4, 0 wrteei 0 # Disable interrupts for rest of handler # The following line is for MPC551x: # stw r4, INTC_EOIR_PRC0@l(r3)# MPC551x: Write 0 to proc'r 0 INTC_EOIR # The following line is for MPC555x: stw r4, INTC_EOIR@l(r3)# MPC555x: Write 0 to INTC_EOIR lwz r3, 0x0C (r1) # Restore SRR0 mtsrr0 r3 lwz r3, 0x10 (r1) # Restore SRR1 mtsrr1 r3 lwz r4, 0x2C (r1) # Restore working registers lwz r3, 0x28 (r1) addi r1, r1, 0x50 # Delete stack frame rfi # End of Interrupt Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 45 5.3.3 # # # # # # intc_hw_branch_table.s (partial) file (MPC555x shown) intc_hw_branch_table.s - INTC hardware vector mode branch table example Description: INTC vector branch table when using INTC in HW vector mode **** NOTE **** ONLY 100 EXAMPLE VECTORS ARE IMPLEMENTED HERE Rev 1.0 Jul 2, 2007 S Mihalik Rev 1.1 Aug 30 1007 SM - Made SwIrq4Handler, emiosCh0Handler .extern Copyright Freescale Semiconductor, Inc. 2007. All rights reserved .section .intc_hw_branch_table .extern SwIrq4Handler .extern emiosCh0Handler #.equ ALIGN_OFFSET, 4 # MPC551x: 4 byte branch alignments (Diab, GHS use 2, CodeWarrior 4) .equ ALIGN_OFFSET, 16 # MPC555x: 16 byte branch alignments (Diab, GHS use 4, CodeWarrior 16) IntcHandlerBranchTable: # Only 100 example vectors are implemented here # MPC555x: This table must have 64 KB alignment # MPC551x: Requires 2 KB alignment after 4KB boundary .align ALIGN_OFFSET hw_vect0: b hw_vect0 #INTC HW vector 0 .align ALIGN_OFFSET hw_vect1: b hw_vect1 #INTC HW vector 1 .align ALIGN_OFFSET hw_vect2: b hw_vect2 #INTC HW vector 2 .align ALIGN_OFFSET hw_vect3: b hw_vect3 #INTC HW vector 3 .align ALIGN_OFFSET hw_vect4: b SwIrq4Handler # SW IRQ 4 .align ALIGN_OFFSET hw_vect5: b hw_vect5 #INTC HW vector 5 ... etc. for other vectors hw_vect50: b hw_vect50 #INTC HW vector 50 .align ALIGN_OFFSET # Use 1 of the next 2 lines hw_vect51: b emiosCh0Handler #MPC555x: eMIOS Ch 0 #hw_vect51: b hw_vect51 #INTC HW vector 51 .align ALIGN_OFFSET hw_vect52: b hw_vect52 #INTC HW vector 52 .align ALIGN_OFFSET hw_vect53: b hw_vect53 #INTC HW vector 53 .align ALIGN_OFFSET hw_vect54: b hw_vect54 #INTC HW vector 54 .align ALIGN_OFFSET hw_vect55: b hw_vect55 #INTC HW vector 55 .align ALIGN_OFFSET hw_vect56: b hw_vect56 #INTC HW vector 56 .align ALIGN_OFFSET hw_vect57: b hw_vect57 #INTC HW vector 57 .align ALIGN_OFFSET # Use 1 of the next 2 lines #hw_vect58: b emiosCh0Handler #MPC551x: eMIOS Ch 0 hw_vect58: b hw_vect58 #INTC HW vector 58 .align ALIGN_OFFSET hw_vect59: b hw_vect59 #INTC HW vector 59 .align ALIGN_OFFSET ... etc. for other vectors Qorivva Simple Cookbook, Rev. 4 46 Freescale Semiconductor 5.3.4 # # # # ivor_branch_table.s file (MPC551x only) ivor_branch_table.s - for use with MPC551x only Description: Branch table for 16 MPC551x core interrupts Copyright Freescale 2007. All Rights Reserved Rev 1.0 Jul 6 2007 S Mihalik - Initial version .section .ivor_branch_table .equ SIXTEEN_BYTES, 16 # 16 byte alignment required for table entries # Diab compiler uses value of 4 (2**4=16) # CodeWarrior, GHS, Cygnus use 16 .align SIXTEEN_BYTES b IVOR0trap # IVOR 0 interrupt handler .align SIXTEEN_BYTES IVOR1trap: b IVOR1trap # IVOR 1 interrupt handler .align SIXTEEN_BYTES IVOR2trap: b IVOR2trap # IVOR 2 interrupt handler .align SIXTEEN_BYTES IVOR3trap: b IVOR3trap # IVOR 3 interrupt handler .align SIXTEEN_BYTES IVOR4trap: b IVOR4trap # IVOR 4 interrupt handler .align SIXTEEN_BYTES IVOR5trap: b IVOR5trap # IVOR 5 interrupt handler .align SIXTEEN_BYTES IVOR6trap: b IVOR6trap # IVOR 6 interrupt handler .align SIXTEEN_BYTES IVOR7trap: b IVOR7trap # IVOR 7 interrupt handler .align SIXTEEN_BYTES IVOR8trap: b IVOR8trap # IVOR 8 interrupt handler .align SIXTEEN_BYTES IVOR9trap: b IVOR9trap # IVOR 9 interrupt handler .align SIXTEEN_BYTES IVOR10trap: b IVORtrap # IVOR 10 interrupt handler .align SIXTEEN_BYTES IVOR11trap: b IVOR11trap # IVOR 11 interrupt handler .align SIXTEEN_BYTES IVOR12trap: b IVOR12trap # IVOR 12 interrupt handler .align SIXTEEN_BYTES IVOR13trap: b IVOR13trap # IVOR 13 interrupt handler .align SIXTEEN_BYTES IVOR14trap: b IVOR14trap # IVOR 14 interrupt handler .align SIXTEEN_BYTES IVOR15trap: b IVOR15trap # IVOR15 interrupt handler IVOR0trap: Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 47 6 INTC: Software Vector Mode, VLE Instructions 6.1 Description Task: Using the Interrupt Controller (INTC) in software vector mode, this program provides two interrupts that show nesting. The main differences from the other INTC SW Mode, Classic Instructions example, are: • VLE instructions are used (not available in MPC5553, MPC5554) • Prologue/epilogue uses VLE assembly instructions • Programmable Interrupt Timer (PIT) is used as the interrupt timer (note: PIT is not available in some MPC55xx devices, but an eMIOS channel could be used) • The timer will count at the system clock rate (causing an interrupt at a count value of 0) • System clock is set to 64 MHz A relative interrupt response can be measured by reading the PIT count value in the first line of the interrupt service routine. An alternate method is to put a breakpoint in the beginning of the service routine and to read the count register. NOTE: to get a true interrupt performance measurement, additional software is needed to initialize branch target buffers, configure flash, and enable cache (if implemented), as shown in other examples in this application note. The interrupt handler will re-enable interrupts and later, in its interrupt service routine (ISR), invoke a second interrupt every other time. This will provide an approximate 1 ms task (ISR) from the PIT and an approximate 2 ms task (ISR) from the software interrupt. The software interrupt will have a higher priority, so the 1 PIT ISR is preempted by the software interrupt. Both ISRs will have a counter. The ISRs will be written in C, so the appropriate registers will be saved and restored in the handler. The SPE will not be used, so its accumulator will not be saved in the stack frame of the interrupt handler. Exercise: Write a third interrupt handler which uses a software interrupt of a higher priority than the others. Invoke this new software interrupt from one of the other ISRs. Crystal 8 MHz Clocks and PLL SIU_SYSCLK (MPC551x only) Divide sysclk by 1 for PIT PIT PIT 1 Timer Load Value PIT 1 Timer sysclk set to 64 MHz MPC5500 / MPC5600 Interrupt Request to INTC Software settable interrupt request 4 (set in PIT interrupt service routine) Interrupt Request to INTC INTC (Software Vector Mode) Common INTC interrupt request to CPU core Figure 13. Software Vector Mode, VLE Instructions Example Qorivva Simple Cookbook, Rev. 4 48 Freescale Semiconductor 6.2 Design The overall program flow for MPC551x and MPC56xxB/P/S is shown below. When an interrupt occurs, it is routed to either or both cores, as defined in the INTC_PSR for that vector. For MPC551x, in this example only one processor is selected for interrupts: processor 0 (e200z1). The selection is defined in INTC_PSR for each enabled interrupt — in this case PIT 1 and software interrupt 4. For MPC551x, there can be a different ISR vector table for each core because each core has its own special purpose register, IVPR. The Vector Table Base Address is defined in each core’s INTC_ACKR, that is, either INTC_ACK_PRC0 (for e200z1) or INTC_IACK_PRC1 (for e200z0). This example only uses the e200z1 core. External Input interrupt taken vector to IVPR0:19 + 4(for IVOR4)16 bytes main.c file ivor_branch_table_vle.s file IVOR Branch Table e_b IVOR0trap e_b IVOR1trap ... e_b IVOR4Handler ... IntcIsrVectors.c file ISR Vector Table: ... SW IRQ 4 ISR addr ... PIT 1 ISR addr ... ... ... handlers_VLE.s or handlers_new_vle.s file IVOR4Handler: prologue: save SRR0:1, fetch ISR vector, re-enable interrupts save registers branch to fetched ISR vector epilogue (return from ISR): restore most registers ensure prior stores completed disable interrupts write to EOIR to restore priority restore SRR0:1 return main { init Modes (MPC56xxBPS only) init IVPR, IVOR4 (MPC551x only) init INTC init PIT1 init SW 4 interrupt enable interrupts wait forever } Pit1ISR { increment counter if odd count, invoke SW 4 IRQ clear flag } SwIrq4ISR { increment counter clear flag } Figure 14. MPC551x, MPC56xxPBS Program Flow Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 49 The overall program flow for MPC555x is shown below. External Input interrupt taken handlers_vle.s or handlers_new_vle.s file IntcIsrVectors.c file ISR Vector Table: ... SW IRQ 4 ISR addr ... PIT 1 ISR addr ... ... ... vector to IVPR0:15 + IVOR416:27 IVOR4Handler: prologue: save SRR0:1, fetch ISR vector, re-enable interrupts save registers branch to fetched ISR vector epilogue (return from ISR): restore most registers ensure prior stores completed disable interrupts write to EOIR to restore priority restore SRR0:1 return main.c file main { init IVPR, IVOR4 init INTC init PIT1 init SW 4 interrupt enable interrupts wait forever } Pit1ISR { increment counter if odd count, invoke SW 4 IRQ clear flag } SwIrq4ISR { increment counter clear flag } Figure 15. MPC555x Program Flow Qorivva Simple Cookbook, Rev. 4 50 Freescale Semiconductor 6.2.1 Modes Use Summary (MPC56xxB/P/S only) Mode Transition is required for changing mode entry registers. Hence even enabling the crystal oscillator to be active in the default mode (DRUN) requires enabling the crystal oscillator in appropriate mode configuration register (ME_xxxx_MC) then initiating a mode transition. After reset, the mode is switched in this example from the default mode (DRUN) to RUN0 mode. The following table summarizes the mode settings used. Table 21. Mode Configurations for MPC56xxB/P/S INTC SW Mode, VLE Instructions Example Modes are enabled in ME_ME Register. Settings Mode Mode Config. Register Value sysclk Selection DRUN ME_DRUN_MC 0x001F 0010 (default) RUN0 Memory Power Mode Clock Sources Mode Config. Register ME_RUN0_MC 0x001F 007D PLL1 16MHz XOSC0 PLL0 (MPC IRC 56xxP/S only) Data Flash Code Flash Main Voltage Reg. I/O Power Down Ctrl 16 MHz IRC On Off Off Off Normal Normal On Off PLL0 On On On Off Normal Normal On Off Other modes are not used in example Peripherals also have configurations to gate clocks on and off, enabling low power. The following table summarizes the peripheral configurations used in this example. Table 22. Peripheral Configurations for MPC56xxB/P/S INTC SW Vector Mode, VLE Example Low power modes are not used in example. PeriPeri. Enabled Modes pheral Config. Config. Register RUN3 RUN2 RUN1 RUN0 DRUN SAFE TEST RESET PC1 ME_ RUNPC_ 1 0 0 0 1 0 0 0 0 Peripherals Selecting Configuration Peripheral PIT, RTI PCTL Reg. # 92 Other peripheral configurations are not used in example Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 51 Table 23. Initialization: INTC in Software Vector Mode, VLE Instructions Pseudo Code Step Relevant Bit Fields MPC551x Variable Init MPC555x Counter for PIT 1 interrupts Counter for software 4 interrupts int Pit1Ctr = 0 int SWirq4Ctr = 0 Table Init Load INTC’s ISR vector table with addresses for: • SWirq4ISR: INTC vector #4 • Pit1ISR: INTC vector: MPC551x: #149, MPC563x: #302, MPC56xxB/P/S: #60 &SwIrq4ISR &Pit1ISR init Enable desired modes DRUN=1, RUN0 = 1 Modes And Clock Initialize PLL0 dividers to provide 64 MHz for input crystal before mode configuration: (MPC • 8 MHz xtal: FMPLL[0]_CR=0x02400100 56xxPBS • 40 MHz xtal: FMPLL[0]_CR=0x12400100 only) (MPC56xxP & 8 MHz xtal requires changing default CMU_CSR value. See PLL example.) Configure RUN0 Mode: • I/O Output power-down: no safe gating (default) • Main Voltage regulator is on (default) • Data, code flash in normal mode (default) • PLL0 is switched on • Crystal oscillator is switched on • 16 MHz IRC is switched on (default) • Select PLL0 (system pll) as sysclk MPC56xxB/S: • Peri. Config.1: run in DRUN mode only - ME_ME = 0x0000 001D - 8 MHz Crystal: CGM_ FMPLL[0]_CR =0x02400100 - ME_ RUN0_MC = 0x001F 0070 - ME_RUN_PC1 = 0000 0010 - ME_PCTL92 = 0x01 - ME_MCTL =0x4000 5AF0, =0x4000 A50F wait ME_GS [S_TRANS] = 0 verify 4 = ME_GS [CURRENTMODE] (See example PLL: Initializing System Clock) - - See PLL Initialization example PDO=0 MVRON=1 DFLAON, CFLAON=3 PLL0ON=1 XOSC0ON=1 16MHz_IRCON=1 SYSCLK=0x4 RUN0 = 1 Assign peripheral configuration to peripherals: • PIT, RTI: select ME_RUN_PC1 RUN_CFG = 0 Initiate software mode transition to RUN0 mode • Mode & key, then mode & inverted key • Wait for transition to complete TARGET_MODE = RUN0 S_TRANS • Verify current mode is RUN0 CURRENTMODE init Sysclk Initialize sysclk to 64 MHz, running from PLL disable Watchdog MPC56xxBPS Disable watchdog by writing keys to Status Register, then clearing WEN (MPC56xxBPS only) Qorivva Simple Cookbook, Rev. 4 52 Freescale Semiconductor Table 23. Initialization: INTC in Software Vector Mode, VLE Instructions (continued) Pseudo Code Step Relevant Bit Fields MPC551x init Irq Vectors init INTC init PIT Load the common interrupt vector prefix to spr IVPR. MPC555x MPC56xxBPS spr IVPR =__IVPR_VALUE MPC555x: Initialize external input interrupt offset to the lower half of IVROR4 handler’s address - Initialize INTC: • Configure for software vector mode (HVEN=0) • Keep vector offset (VTES) at 4 bytes (MPC551x note: only proc’r 0, e200z1 is used) HVEN_PRC0=0(551x) VTES_PRC0=0(551x) HVEN = 0 (MPC555x) VTES = 0 (MPC555x) Initialize ISR vector table base address per link file (MPC551x note: only proc’r 0, e200z1, is used here) VTBA = passed from linker spr IVOR4 = IVOR4Handler @l - INTC_MCR = 0x0 INTC_IACKR _PRC 0 = __IACKR_ VTBA_VALUE INTC_IACKR = __IACKR_ VTBA_VALUE MPC551x: Init system clock divider for PIT, LPCLKDIV1 = 0 RTI to divide by 1 (also applies to other Group (default) 1 peripherals of eSCI A, IIC) SIU_SYSCLK [LPCLKDIV1] =0 - Global controls: • Enable module • MPC555x, 56xxBPS: Freeze in debug mode PIT_CTRL = 0x0000 0000 MDIS = 0 FRZ = 1 Load a start count values for 64 MHz sysclk (PITs count down from value at sysclk rate) • PIT 1 Timeout = 64M/(64M sysclk/sec)=1ms PIT_TVAL1 = 64,000 PIT_PITMCR = 0x0000 0001 PIT_TIMER1_ LDVAL1 = 64,000 PIT_PITEN = PEN1(551x) = 1 or TEN(555x,56xxPBS)=1 0x0000 0002 PIT_TIMER1_ TIE1 or TIE = 1 PIT_INTEN = TCTRL = Enable PIT 1 to request IRQ 0x0000 0002 0x00000003 PIT_INTSEL = MPC551x: Assign PIT 1 flag for IRQ instead ISEL = 1 0x0000 0002 of DMA PIT_CH1_ LDVAL = 64,000 Enable PIT 1 timer to count (counts down) Configure PIT1 interrupts: • MPC551x: Select processor 0 (e200z1) • Select priority 1 PRC_SEL= 0 (z1) PRI = 1 init SwIrq4 Software Interrupt 4: • Raise priority to 2 PRI = 2 • MPC551x: Select processor(s) to interrupt PRC_SEL=0 (e200z1) enable ExtIrq Enable recognition of requests to INTC by PRI = 0 lowering the INTC’s current priority to 0 from default of 15 waitloop INTC_PSR [149] = 0x01 INTC_PSR [302] = 0x01 PIT_CH1_ TCTRL = 0x0000 0003 INTC_PSR [60] = 0x01 INTC_PSR[4] = 0x02 INTC_CPR _PRC0[PRI]= 0 INTC_CPR [PRI]= 0 Enable external interrupts: set MSR[EE] = 1 wrteei 1 wait forever while 1 Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 53 6.2.2 Interrupt Handler Table 24. Stack Frame for INTC Interrupt Handler Stack Frame Area 32 bit GPRs Register Location in Stack r12 sp + 0x4C r11 sp + 0x48 r10 sp + 0x44 r9 sp + 0x40 r8 sp + 0x3C r7 sp + 0x38 r6 sp + 0x34 r5 sp + 0x30 r4 sp + 0x2C r3 sp + 0x28 r0 sp + 0x24 CR CR sp + 0x20 locals and padding XER sp + 0x1C CTR sp + 0x18 LR sp + 0x14 SRR1 sp + 0x10 SRR0 sp + 0x0C padding sp + 0x08 LR area LR placeholder sp + 0x04 back chain r1 sp Qorivva Simple Cookbook, Rev. 4 54 Freescale Semiconductor Table 25. IVOR4 Interrupt Handler (INTC in Software Vector Mode, VLE Instructions) (MPC555x: Must be 16-byte aligned and within 64 KB of address in IVPR. MPC551x does not require alignment because it is the destination of a branch instruction.) Step prolog Pseudo Code Relevant Bit Fields Create stack frame e_stwu sp, -0x50 (sp) Save SRR0:1 because nested interrupts will be allowed store SRR0:1, r3 registers to stack frame Read pointer into ISR Vector Table Re-enable external interrupts by setting MSR[EE] r3 = INTC_IACKR_PRC0 EE = 1 se_lwz r3, 0x0 (r3) mtLR r3 Save other appropriate registers for C ISR store other registers to stack frame Branch to ISR, saving return address se_blrl Restore registers from stack frame except SRR0:1 and two working registers load most registers from stack frame Ensure interrupt flag in ISR has completed clearing before writing to INTC_EOIR Disable external interrupts by clearing MSR[EE] Restore former INTC’s Current Priority Restore SRR0:1 and working registers r3 = INTC_IACKR wrteei 1 Read ISR address from ISR Vector Table and store into LR epilog MPC555x MPC56xxBPS MPC551x mbar EE = 0 wrteei 0 write 0 to INTC_EOIR_PRC0 write 0 to INTC_EOIR restore SRR0:1, working registers from stack frame Restore stack frame space e_add16i sp, sp, 0x50 Return se_rfi Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 55 6.2.3 Interrupt Service Routines (ISRs) Table 26. ISR for PIT1 Relevant Bit Fields Step Pseudo Code MPC551x MPC555x Pit1ISR Increment Pit1Ctr MPC56xxB/P/S Pit1Ctr ++ If Pit1Ctr is even, invoke Software Interrupt 4 SET=1 if ((Pit1Ctr&1) ==0), INTC_SSCIR[4] = 2 Clear Pit1Ctr interrupt flag by writing 1 to it TIF1 = 1 or PIT_PITFLG[TIF1] PIT_TFLG1[TIF] PIT_CH1_TFLG[TIF] TIF = 1 =1 =1 =1 Table 27. ISR for Software Interrupt 4 Step Relevant Bit Fields swIRQ4ISR Increment SWirq4ctr Clear Software IRQ 4 Pseudo Code SWirq4ctr++ CLR=1 INTC_SSCIR[4] = 1 Qorivva Simple Cookbook, Rev. 4 56 Freescale Semiconductor 6.3 6.3.1 6.3.1.1 /* /* /* /* Code main.c files main.c (MPC551x shown with 8 MHz crystal) main.c - Software vector mode program using C isr */ Feb 03 2009 SM - Based on AN2865 INTC Software Vector Mode example*/ Aug 12 2009 SM - Changed initial ERFD value, added 12MHz crystal numbers */ Copyright Freescale Semiconductor, In.c 2009 All rights reserved. */ #include "mpc5510.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ extern IVOR4Handler(); extern uint32_t __IVPR_VALUE; /* Interrupt Vector Prefix value from link file*/ extern const vuint32_t IntcIsrVectorTable[]; uint32_t Pit1Ctr = 0; /* Counter for PIT 1 interrupts */ uint32_t SWirq4Ctr = 0;/* Counter for software interrupt 4 */ void initSysclk(void) { /* Use 2 of the next 4 lines: */ FMPLL.ESYNCR2.R = 0x00000007; /* 8MHz xtal: ERFD to initial value of 7 */ FMPLL.ESYNCR1.R = 0xF0000020; /* 8MHz xtal: CLKCFG=PLL, EPREDIV=0, EMFD=0x20 */ /*FMPLL.ESYNCR2.R = 0x00000005; */ /* 12MHz xtal: ERFD to initial value of 5 */ /*FMPLL.ESYNCR1.R = 0xF0020030; */ /* 12MHz xtal: CLKCFG=PLL, EPREDIV=2, EMFD=0x30*/ CRP.CLKSRC.B.XOSCEN = 1; /* Enable external oscillator */ while (FMPLL.SYNSR.B.LOCK != 1) {}; /* Wait for PLL to LOCK */ /* Use 1 of the next 2 lines: */ FMPLL.ESYNCR2.R = 0x00000005; /* 8MHz xtal: ERFD change for 64 MHz sysclk */ /*FMPLL.ESYNCR2.R = 0x00000003; */ /* 12MHz xtal: ERFD change for 64 MHz sysclk */ SIU.SYSCLK.B.SYSCLKSEL = 2; /* Select PLL for sysclk */ } void disableWatchdog(void) { SWT.SR.R = 0x0000c520; /* Write keys to clear soft lock bit */ SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; /* Clear watchdog enable (WEN) */ } asm void initIrqVectors(void) { lis r3, __IVPR_VALUE@h ori r3, r3, __IVPR_VALUE@l mtivpr r3 } /* IVPR value is passed from link file */ void initINTC(void) { INTC.MCR.B.HVEN_PRC0 = 0; /* MPC551x Proc'r 0: initialize for SW vector mode*/ INTC.MCR.B.VTES_PRC0 = 0; /* MPC551x Proc'r 0: default vector table 4B offsets */ INTC.IACKR_PRC0.R = (uint32_t) &IntcIsrVectorTable[0]; /* MPC551x: ISR table base*/ } void initPIT(void) { SIU.SYSCLK.B.LPCLKDIV1 = 0; PIT.PITCTRL.R = 0; PIT.TLVAL[1].R = 64000; PIT.PITINTEN.R = 0x00000002; PIT.PITINTSEL.R = 0x00000006; PIT.PITEN.B.PEN1 = 1; INTC.PSR[149].R = 0x01; } /* /* /* /* /* /* /* Divide sysclk by 1 for Group 1 modules */ Ensure PIT is not disbaled */ Timeout= 64K sysclks x 1sec/64M sysclks= 1 ms */ Enable PIT 1 flag to request INTC or DMA */ Assign PIT 1 flag to select IRQ, not DMA req. */ Start PIT counting */ PIT 1 interrupt selects proc'r 0 & priority 1 */ void initSwIrq4(void) { INTC.PSR[4].R = 2;/* Software interrupt 4 IRQ priority = 2 */ } Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 57 void enableIrq(void) { INTC.CPR_PRC0.B.PRI = 0; asm(" wrteei 1"); } /* MPC551x Proc'r 0: Lower INTC's current priority */ /* Enable external interrupts */ void main (void) { vuint32_t i = 0; /* Dummy idle counter */ } initSysclk(); initIrqVectors(); initINTC(); initPIT(); initSwIrq4(); enableIrq(); while (1) { i++; } void Pit1ISR(void) { Pit1Ctr++; if ((Pit1Ctr & 1)==0) { INTC.SSCIR[4].R = 2; } PIT.PITFLG.B.TIF1 = 1; } void SwIrq4ISR(void) { SWirq4Ctr++; INTC.SSCIR[4].R = 1; } /* /* /* /* /* /* Set sysclk to 64 MHz */ Initialize exceptions: only need to load IVPR */ Initialize INTC for software vector mode */ Initialize PIT1 for 1kHz IRQ, priority 2 */ Initialize software interrupt 4 */ Ensure INTC current prority=0 & enable IRQ */ /* Increment interrupt counter */ /* If PIT1Ctr is even*/ /* then invoke software interrupt 4 */ /* Clear PIT 1 flag by writing 1 */ /* Increment interrupt counter */ /* Clear channel's flag */ Qorivva Simple Cookbook, Rev. 4 58 Freescale Semiconductor 6.3.1.2 main.c (MPC555x: MPC563xM shown with 8 MHz crystal) /* main.c - Software vector mode program using C isr */ /* Feb 03 2009 SM - Based on AN2865 INTC Software Vector Mode example*/ /* Copyright Freescale Semiconductor, In.c 2009 All rights reserved. */ #include "mpc563m.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ extern IVOR4Handler(); extern uint32_t __IVPR_VALUE; /* Interrupt Vector Prefix value from link file*/ extern const vuint32_t IntcIsrVectorTable[]; uint32_t Pit1Ctr = 0; /* Counter for PIT 1 interrupts */ uint32_t SWirq4Ctr = 0; /* Counter for software interrupt 4 */ void initSysclk(void) { /* Intialize sysclk to 64MHz for 8 MHz crystal*/ FMPLL.ESYNCR1.B.CLKCFG = 0X7; /* Change clk to PLL normal mode from crystal */ FMPLL.SYNCR.R = 0x16080000; /* Initial values: PREDIV=1, MFD=12, RFD=1 */ while (FMPLL.SYNSR.B.LOCK != 1) {}; /* Wait for FMPLL to LOCK */ FMPLL.SYNCR.R = 0x16000000; /* Final value for 64 MHz: RFD=0 */ } asm void initIrqVectors(void) { lis r3, __IVPR_VALUE@h /* IVPR value is passed from link file */ ori r3, r3, __IVPR_VALUE@l mtivpr r3 li r3, IVOR4Handler@l /* IVOR4 = lower half of handler address */ mtivor4r3 } void initINTC(void) { INTC.MCR.B.HVEN = 0; /* Initialize for SW vector mode */ INTC.MCR.B.VTES = 0; /* Use default vector table 4B offsets */ INTC.IACKR.R=(uint32_t) &IntcIsrVectorTable[0]; /* INTC ISR table base */ } void initPIT(void) { PIT.MCR.R = 0x00000001; /* Enable PIT module & freeze count during debug */ PIT.TIMER[1].LDVAL.R = 64000; /* Timeout= 64K sysclks x 1sec/64M sysclks= 1 ms */ PIT.TIMER[1].TCTRL.R = 0x00000003; /* Start timer counting & freeze during debug */ INTC.PSR[302].R = 0x1; /* PIT 1 interrupt has priority 1 */ } void initSwIrq4(void) { INTC.PSR[4].R = 2; /* Software interrupt 4 IRQ priority = 2 */ } void enableIrq(void) { INTC.CPR.B.PRI = 0; /* Lower INTC's current priority */ asm(" wrteei 1"); /* Enable external interrupts */ } void main (void) { vuint32_t i = 0; } initSysclk(); initIrqVectors(); initINTC(); initPIT(); initSwIrq4(); enableIrq(); while (1) { i++; } /* Dummy idle counter */ /* /* /* /* /* /* Set sysclk to 64 MHz */ Initialize exceptions: only need to load IVPR */ Initialize INTC for software vector mode */ Initialize PIT1 for 1kHz IRQ, priority 2 */ Initialize software interrupt 4 */ Ensure INTC current prority=0 & enable IRQ */ void Pit1ISR(void) { Pit1Ctr++; /* Increment interrupt counter */ if ((Pit1Ctr & 1)==0) { /* If PIT1Ctr is even*/ INTC.SSCIR[4].R = 2; /* then nvoke software interrupt 4 */ } PIT.TIMER[1].TFLG.B.TIF = 1; /* Clear PIT 1 flag by writing 1 */ } void SwIrq4ISR(void) { SWirq4Ctr++; /* Increment interrupt counter */ INTC.SSCIR[4].R = 1; /* Clear channel's flag */ } Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 59 6.3.1.3 /* /* /* /* /* /* main.c file (MPC56xxB/P/S - MPC56xxS shown with 8 MHz crystal) main.c - Software vector mode program using C isr */ Jan 15, 2009 S.Mihalik- Initial version based on previous AN2865 example */ May 22 2009 S. Mihalik- Simplifed by removing unneeded sysclk, PCTL code */ Jul 03 2009 S Mihalik - Simplified code */ Mar 15 2010 SM - modified initModesAndClks, updated header */ Copyright Freescale Semiconductor, Inc. 2009, 2010. All rights reserved. */ #include "56xxS_0204.h" /* Use proper include file */ extern IVOR4Handler(); extern uint32_t __IVPR_VALUE; /* Interrupt Vector Prefix vaue from link file*/ extern const vuint32_t IntcIsrVectorTable[]; uint32_t Pit1Ctr = 0; /* Counter for PIT 1 interrupts */ uint32_t SWirq4Ctr = 0; /* Counter for software interrupt 4 */ void initModesAndClock(void) { ME.MER.R = 0x0000001D; } /* Enable DRUN, RUN0, SAFE, RESET modes */ /* Initialize PLL before turning it on: */ CGM.FMPLL[0].CR.R = 0x02400100; /* 8 MHz xtal: Set PLL0 to 64 MHz */ ME.RUN[0].R = 0x001F0074; /* RUN0 cfg: 16MHzIRCON,OSC0ON,PLL0ON,syclk=PLL */ ME.RUNPC[1].R = 0x00000010; /* Peri. Cfg. 1 settings: only run in RUN0 mode */ ME.PCTL[92].R = 0x01; /* PIT, RTI: select ME_RUN_PC[1] */ /* Mode Transition to enter RUN0 mode: */ ME.MCTL.R = 0x40005AF0; /* Enter RUN0 Mode & Key */ ME.MCTL.R = 0x4000A50F; /* Enter RUN0 Mode & Inverted Key */ while (ME.GS.B.S_MTRANS) {} /* Wait for mode transition to complete */ /* Note: could wait here using timer and/or I_TC IRQ */ while(ME.GS.B.S_CURRENTMODE != 4) {} /* Verify RUN0 is the current mode */ void disableWatchdog(void) { SWT.SR.R = 0x0000c520; /* Write keys to clear soft lock bit */ SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; /* Clear watchdog enable (WEN) */ } asm void initIrqVectors(void) { lis r3, __IVPR_VALUE@h /* IVPR value is passed from link file */ ori r3, r3, __IVPR_VALUE@l mtivpr r3 } void initINTC(void) { INTC.MCR.B.HVEN = 0; /* Initialize for SW vector mode */ INTC.MCR.B.VTES = 0; /* Use default vector table 4B offsets */ INTC.IACKR.R = (uint32_t) &IntcIsrVectorTable[0]; /* INTC ISR table base */ } void initPIT(void) { PIT.MCR.R = 0x00000001; /* Enable PIT and configure stop in debug mode */ PIT.CH[1].LDVAL.R = 64000; /* Timeout= 64K sysclks x 1sec/64M sysclks= 1 ms */ PIT.CH[1].TCTRL.R = 0x000000003; /* Enable PIT1 interrupt & start PIT counting */ INTC.PSR[60].R = 0x01; /* PIT 1 interrupt vector with priority 1 */ } void initSwIrq4(void) { INTC.PSR[4].R = 2; } /* Software interrupt 4 IRQ priority = 2 */ void enableIrq(void) { INTC.CPR.B.PRI = 0; asm(" wrteei 1"); } /* Lower INTC's current priority */ /* Enable external interrupts */ Qorivva Simple Cookbook, Rev. 4 60 Freescale Semiconductor void main (void) { vuint32_t i = 0; initModesAndClock(); disableWatchdog(); initIrqVectors(); initINTC(); initPIT(); initSwIrq4(); enableIrq(); } /* Dummy idle counter */ /* /* /* /* /* /* /* MPC56xxP/B/S: Initialize mode entries, set sysclk=64 MHz*/ Disable watchdog */ Initialize exceptions: only need to load IVPR */ Initialize INTC for software vector mode */ Initialize PIT1 for 1kHz IRQ, priority 2 */ Initialize software interrupt 4 */ Ensure INTC current prority=0 & enable IRQ */ while (1) { i++; } void Pit1ISR(void) { Pit1Ctr++; if ((Pit1Ctr & 1)==0) { INTC.SSCIR[4].R = 2; } PIT.CH[1].TFLG.B.TIF = 1; } void SwIrq4ISR(void) { SWirq4Ctr++; INTC.SSCIR[4].R = 1; } /* Increment interrupt counter */ /* If PIT1Ctr is even*/ /* then nvoke software interrupt 4 */ /* CLear PIT 1 flag by writing 1 */ /* Increment interrupt counter */ /* Clear channel's flag */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 61 6.3.2 # # # # # # # # # handlers_vle.s file handlers_vle.s - INTC software vector mode example using VLE instructions Description: Creates prolog, epilog for C ISR and enables nested interrupts Rev 1.0: April 23, 2004, S Mihalik, Rev 1.1 Aug 2, 2004 SM - delayed writing to EOIR until after disabling EE in epilog Rev 1.2 Sept 8 2004 SM - optimized & corrected r3,r4 restore sequence from rev 1.1 Rev 1.2 Sept 21 2004 SM - optimized by minimizing time interrupts are disabled Rev 1.3 Jul 2 2007 SM - Changes for MPC551x and mapped to .ivor_handlers section Rev 1.4 Jan 22 2009 SM - Modified for VLE instructions, CodeWarrior 2.4 alpha Copyright Freescale Semiconductor, Inc. 2007. All rights reserved # STACK FRAME DESIGN: Depth: 20 words (0xA0, or\ 80 bytes) # ************* ______________ # 0x4C * GPR12 * ^ # 0x48 * GPR11 * | # 0x44 * GPR10 * | # 0x40 * GPR9 * | # 0x3C * GPR8 * | # 0x38 * GPR7 * GPRs (32 bit) # 0x34 * GPR6 * | # 0x30 * GPR5 * | # 0x2C * GPR4 * | # 0x28 * GPR3 * | # 0x24 * GPR0 * ___v__________ # 0x20 * CR * __CR__________ # 0x1C * XER * ^ # 0x18 * CTR * | # 0x14 * LR * locals & padding for 16 B alignment # 0x10 * SRR1 * | # 0x0C * SRR0 * | # 0x08 * padding * ___v__________ # 0x04 * resvd- LR * Reserved for calling function # 0x00 * SP * Backchain (same as gpr1 in GPRs) # ************* .section .ivor_handlers,text_vle .globlIVOR4Handler .align 16 # Align IVOR handlers on a 16 byte boundary for MPC555x # GHS, Cygnus, Diab(default) use .align 4; Metrowerks .align 16 .equINTC_IACKR_PRC0, 0xfff48010 .equINTC_EOIR_PRC0, 0xfff48018 .equINTC_IACKR, 0xfff48010 .equINTC_EOIR,0xfff48018 # # # # MPC551x: Proc 0 Interrupt Acknowledge Reg. addr. MPC551x: Proc 0 End Of Interrupt Reg. addr. Single Core: Interrupt AcknowledgeReg. addr. Single Core: End Of Intterrupt Reg. addr. IVOR4Handler: prolog: e_stwu se_stw mfsrr0 se_stw mfsrr1 se_stw e_lis e_lwz # e_lis # e_lwz r1, -0x50 (r1) r3, 0x28 (r1) r3 r3, r3 r3, 0x0C (r1) # PROLOGUE # Create stack frame and store back chain # Store a working register # Note: use se_ form for r0-7, r24-31 with positive offset # Store SRR0:1 (must be done before enabling EE) 0x10 (r1) # Use 2 of the next 4 lines for appropriate processor: r3, INTC_IACKR@ha # Single core: Read pointer into ISR Vector Table r3, INTC_IACKR@l(r3) # Single core r3, INTC_IACKR_PRC0@ha # MPC551x: Read proc'0 pointer into ISR Vector Table r3, INTC_IACKR_PRC0@l(r3) # MPC551x se_lwz r3, 0x0(r3) # Read ISR address from ISR Vector Table using pointer wrteei se_stw se_mflr se_stw 1 r4, r4 r4, # Set MSR[EE]=1(wait a couple clocks after reading IACKR) # Store a second working register # Store LR (LR will be used for ISR Vector) 0x2C (r1) 0x14 (r1) Qorivva Simple Cookbook, Rev. 4 62 Freescale Semiconductor se_mtlr r3 # Store ISR address to LR to use for branching later e_stw r12, 0x4C e_stw r11, 0x48 e_stw r10, 0x44 e_stw r9, 0x40 e_stw r8, 0x3C se_stw r7, 0x38 se_stw r6, 0x34 se_stw r5, 0x30 se_stw r0, 0x24 mfcr r3 se_stw r3, 0x20 mfxer r3 se_stw r3, 0x1C se_mfctr r3 se_stw r3, 0x18 (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) se_blrl # e_lis e_lis # e_stw e_stw # Store XER # Store CTR (r1) # EPILOGUE # Restore LR (r1) # Restore CTR (r1) # Restore XER (r1) # Restore CR (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) # Restore other gprs except working registers 0 # Ensure store to clear interrupt's flag bit completed # Use 1 of the following 2 lines: r3, INTC_EOIR_PRC0@ha # MPC551x: Load upper half of proc'r 0 EIOR address r3, INTC_EOIR@ha # Single Core: Load upper half of EIOR address to r3 se_li wrteei # Store CR # Branch to ISR, but return here epilog: se_lwz r3, 0x14 se_mtlr r3 se_lwz r3, 0x18 se_mtctr r3 se_lwz r3, 0x1C mtxer r3 se_lwz r3, 0x20 mtcrf 0xff, r3 se_lwz r0, 0x24 se_lwz r5, 0x30 se_lwz r6, 0x34 se_lwz r7, 0x38 e_lwz r8, 0x3C e_lwz r9, 0x40 e_lwz r10, 0x44 e_lwz r11, 0x48 e_lwz r12, 0x4C mbar # Store rest of gprs r4, 0 0 # Disable interrupts for rest of handler # Use 1 of the following 2 lines: r4, INTC_EOIR_PRC0@l(r3) # MPC551x - Write 0 to proc'r 0 INTC_EIOR r4, INTC_EOIR@l(r3) # Single Core - Write 0 to proc'r 0 INTC_EIOR se_lwz r3, mtsrr0 r3 se_lwz r3, mtsrr1 r3 se_lwz r4, se_lwz r3, e_add16i r1, 0x0C (r1) # Restore SRR0 0x10 (r1) # Restore SRR1 0x2C (r1) 0x28 (r1) r1, 0x50 # Restore working registers se_rfi # Delete stack frame # End of Interrupt Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 63 6.3.3 /* /* /* /* /* /* /* /* /* /* IntcIsrVectors.c file IntcIsrVectors.c - table of ISRs for INTC in SW vector Mode */ Description: Contains addresses for first 250 ISR vectors */ Table address gets loaded to INTC_IACKR */ Alignment: MPC551x &MPC56xxP/B/S: 2 KB after a 4KB boundary; MPC555x: 64 KB*/ April 22, 2004 S. Mihalik */ March 16, 2006 S. Mihalik - Modified for compile with Diab 5.3 */ Jun 29 2006 SM - Used pragma align instead of hard coding address */ Jul 5 2007 SM - alignment now done in link file; changes for MPC551x */ Aug 30 2007 SM - Added pragma for CodeWarrior */ Oct 22 2008 SM - Changed to use PIT1 ISR instead of eMIOS Ch 0 ISR */ #include "typedefs.h" void dummy (void); extern void SwIrq4ISR(void); extern void Pit1ISR(void); /* Use next two lines with Diab compile */ /*#pragma section CONST ".intc_sw_isr_vector_table" */ /* Diab compiler pragma */ /*const uint32_t IntcIsrVectorTable[] = { */ /* Use pragma next two lines with CodeWarrior compile */ #pragma section data_type ".intc_sw_isr_vector_table" ".intc_sw_isr_vector_table" data_mode=far_abs uint32_t IntcIsrVectorTable[] = { (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&SwIrq4ISR, /* ISRs 00 - 04 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 05 - 09 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 10 - 14 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 15 - 19 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 20 - 24 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 25 - 29 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 30 - 34 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 35 - 39 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 40 - 44 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 45 - 49 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 50 - 54 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 55 - 59 */ /* Use the next line for MPC551x or MPC563x: */ /* (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy,*/ /* ISRs 60 - 64 */ /* Use the next line for MPC56xB, MPC56xxP, MPC56xS, where PIT1 vector number is 60: */ (uint32_t)&Pit1ISR, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 60 - 64 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 65 - 69 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 70 - 74 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 75 - 79 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 80 - 84 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 85 - 89 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 90 - 94 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 95 - 99 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 100 - 104 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 105 - 109 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 110 - 114 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 115 - 119 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 120 - 124 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 125 - 129 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 130 - 134 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 135 - 139 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 140 - 144 */ /* Use the next line for MPC563x or MPC56xxB, MPC56xxP, MPC56xxS: */ /* (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&Pit1ISR,*/ /* ISRs 145 - 149 */ /* Use the next line for MPC551x, where PIT1 vector number is 149: */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 145 - 149 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 150 - 154 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 155 - 159 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 160 - 164 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 155 - 169 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 170 - 174 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 175 - 179 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 180 - 184 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 185 - 189 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 190 - 194 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 195 - 199 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 200 - 204 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 205 - 209 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 210 - 214 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 215 - 219 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 220 - 224 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 225 - 229 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 230 - 234 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 235 - 239 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, /* ISRs 240 - 244 */ (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy, (uint32_t)&dummy /* ISRs 245 - 249 */ /* For MPC563x, continue vectors and add Pit1ISR address for INTC vector number 302 */ }; void dummy (void) { while (1) {}; /* Wait forever or for watchdog timeout */ } Qorivva Simple Cookbook, Rev. 4 64 Freescale Semiconductor 6.3.4 # # # # # # ivor_branch_table.s file (MPC551x, MPC56xxPBS) ivor_branch_table.s - for use with MPC551x, MPC56xxP, MPC56xxB, MPC56xxS only Description: Branch table for 16 MPC551x core interrupts Copyright Freescale 2007. All Rights Reserved Rev 1.0 Jul 6 2007 S Mihalik - Initial version Rev 1.1 Aug 30 2007 SM - Made IVOR4Handler extern Rev 1.2 Sep 9 2008 SM - Converted assembly to VLE syntax .extern IVOR4Handler .section .ivor_branch_table,text_vle .equ SIXTEEN_BYTES, 16 IVOR0trap: IVOR1trap: IVOR2trap: IVOR3trap: IVOR5trap: IVOR6trap: IVOR7trap: IVOR8trap: IVOR9trap: IVOR10trap: IVOR11trap: IVOR12trap: IVOR13trap: IVOR14trap: IVOR15trap: # 16 byte alignment required for table entries # Diab compiler uses value of 4 (2**4=16) # CodeWarrior, GHS, Cygnus use 16 .align SIXTEEN_BYTES e_b IVOR0trap # IVOR 0 interrupt handler .align SIXTEEN_BYTES e_b IVOR1trap # IVOR 1 interrupt handler .align SIXTEEN_BYTES e_b IVOR2trap # IVOR 2 interrupt handler .align SIXTEEN_BYTES e_b IVOR3trap # IVOR 3 interrupt handler .align SIXTEEN_BYTES e_b IVOR4Handler # IVOR 4 interrupt handler (External Interrupt) .align SIXTEEN_BYTES e_b IVOR5trap # IVOR 5 interrupt handler .align SIXTEEN_BYTES e_b IVOR6trap # IVOR 6 interrupt handler .align SIXTEEN_BYTES e_b IVOR7trap # IVOR 7 interrupt handler .align SIXTEEN_BYTES e_b IVOR8trap # IVOR 8 interrupt handler .align SIXTEEN_BYTES e_b IVOR9trap # IVOR 9 interrupt handler .align SIXTEEN_BYTES e_b IVOR10trap # IVOR 10 interrupt handler .align SIXTEEN_BYTES e_b IVOR11trap # IVOR 11 interrupt handler .align SIXTEEN_BYTES e_b IVOR12trap # IVOR 12 interrupt handler .align SIXTEEN_BYTES e_b IVOR13trap # IVOR 13 interrupt handler .align SIXTEEN_BYTES e_b IVOR14trap # IVOR 14 interrupt handler .align SIXTEEN_BYTES e_b IVOR15trap # IVOR15 interrupt handler Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 65 7 INTC: Hardware Vector Mode, VLE Instructions 7.1 Description Task: Using the Interrupt Controller (INTC) in hardware vector mode, this program provides two interrupts that show nesting. The main differences from the other INTC SW Mode, Classic Instructions example, are: • VLE instructions are used (not available on MPC5553, MPC5554) • Prologue/epilogue uses VLE assembly instructions • Programmable Interrupt Timer (PIT) is used as the interrupt timer (note: PIT is not available in some MPC55xx devices, but an eMIOS channel could be used) • Timer will count at the system clock rate (causing an interrupt at a count value of 0) • System clock is set to 64 MHz A relative interrupt response can be measured by reading the PIT count value in the first line of the interrupt service routine. An alternate method is to put a breakpoint in the beginning of the service routine and read the count register. NOTE: to get a true interrupt performance measurement, additional software is needed to initialize branch target buffers, configure flash, and enable cache (if implemented), as shown in other examples in this application note. The interrupt handler will re-enable interrupts and later, in its interrupt service routine (ISR), invoke a second interrupt every other time. This provides an approximate 1 ms task (ISR) from the PIT and an approximate 2 ms task (ISR) from the software interrupt. The software interrupt will have a higher priority, so the 1 PIT ISR is preempted by the software interrupt. Both ISRs will have a counter. The ISRs will be written in C, so the appropriate registers will be saved and restored in the handler. The SPE will not be used, so its accumulator will not be saved in the stack frame of the interrupt handler. Exercise: Write a third interrupt handler which uses a software interrupt of a higher priority than the others. Invoke this new software interrupt from one of the other ISRs. Crystal 8 MHz Clocks and PLL SIU_SYSCLK (MPC551x only) Divide sysclk by 1 for PIT PIT PIT 1 Timer Load Value PIT 1 Timer sysclk set to 64 MHz MPC5500 / MPC5600 Software settable interrupt request 4 (set in PIT interrupt service routine) Interrupt Request to INTC INTC (Hardware Interrupt Vector Request to Mode) INTC Unique INTC interrupt request to CPU core Figure 16. Hardware Vector Mode, VLE Instructions Example Qorivva Simple Cookbook, Rev. 4 66 Freescale Semiconductor 7.2 Design The overall program flow is shown below, followed by design steps and a stack frame implementation. The INTC when used in hardware vector mode uses a branch table for getting to each INTC vector’s handler. These are shown as handler_0, handler_1, etc. (Note: the code in this example does not have handlers for each INTC vector, so a common dummy trap address is used.) For MPC551x, the second core has its own special purpose register IVPR, so the second core would have its own IntcHanderBranchTable (not included in this example). Also for MPC551x, either or both processors can receive the interrupt request from the Interrupt Controller. In this example, only one processor is selected in the MPC551x: processor 0 (e200z1). The selection is defined in INTC_PSR for each enabled interrupt, which here is PIT1 and software interrupt 4. INTC vector for PIT1 interrupt taken vector to: MPC551x: IVPR0:19 + 0x800 + 149 (0x4) MPC555x: IVPR0:15 + 302 (0x10) MPC56xxBPS: IVPR0:19 + 0x800 + 60 (0x4) handlers.s file intc_hw_branch_table.s file IntcHandlerBranchTable b handler_0 ... b handler_1 ... b handler_2 ... b handler_3 ... b SwIrq4Handler ... ... b Pit1Handler ... ... ... main.c file Pit1Handler: prologue: save registers re-enable interrupts branch to Pit1ISR epilogue (return from ISR): restore most registers ensure prior stores completed disable interrupts write to EOIR to restore priority restore SRR0:1 return SwIrq4Handler: prologue: save registers re-enable interrupts branch to SwIrq4ISR epilogue (return from ISR): restore most registers ensure prior stores completed disable interrupts write to EOIR to restore priority restore SRR0:1 return main { init IVPR init INTC init SwIrq4 init PIT1 enable interrupts wait forever } Pit1ISR { increment counter if odd count, invoke SW 4 IRQ clear flag } SwIrq4ISR { increment counter clear flag } Figure 17. INTC HW Vector Mode, VLE Instructions. Overall Program Flow showing PIT 1 Interrupt Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 67 7.2.1 Modes Use Summary (MPC56xxB/P/S only) Mode Transition is required for changing mode entry registers. Hence even enabling the crystal oscillator to be active in the default mode (DRUN) requires enabling the crystal oscillator in appropriate mode configuration register (ME_xxxx_MC) then initiating a mode transition. After reset, the mode is switched in this example from the default mode (DRUN) to RUN0 mode. The following table summarizes the mode settings used. Table 28. Mode Configurations for MPC56xxB/P/S INTC HW Vector Mode VLE Example Modes are enabled in ME_ME Register. Settings Memory Power Mode Clock Sources Mode Mode Config. Register Mode Config. Register Value sysclk Selection DRUN ME_DRUN_MC 0x001F 0010 (default) RUN0 ME_RUN0_MC 0x001F 007D PLL1 16MHz XOSC0 PLL0 (MPC IRC 56xxP/S only) Data Flash Code Flash Main Voltage Reg. I/O Power Down Ctrl 16 MHz IRC On Off Off Off Normal Normal On Off PLL0 On On On Off Normal Normal On Off Other modes are not used in example Peripherals also have configurations to gate clocks on and off, enabling low power. The following table summarizes the peripheral configurations used in this example. Table 29. Peripheral Configurations for MPC56xxB/P/S INTC HW Vector Mode VLE Example Low power modes are not used in example. PeriPeri. Enabled Modes pheral Config. Config. Register RUN3 RUN2 RUN1 RUN0 DRUN SAFE TEST RESET PC1 ME_ RUNPC_ 1 0 0 0 1 0 0 0 0 Peripherals Selecting Configuration Peripheral PIT, RTI PCTL Reg. # 92 Other peripheral configurations are not used in example Qorivva Simple Cookbook, Rev. 4 68 Freescale Semiconductor Initialization Table 30. Initialization: INTC in Hardware Vector Mode, VLE Instructions Pseudo Code Step Relevant Bit Fields MPC551x Variable Init Counter for PIT 1interrupts Counter for software 4 interrupts int Pit1Ctr = 0 int SWirq4Ctr = 0 Table Init Load INTC HW Branch Table with ISR names for: INTC Vector for SW interrupt request 4 INTC Vector for PIT 1 init Enable desired modes Modes And Clock Initialize PLL0 dividers to provide 64 MHz for input crystal before mode configuration: (MPC • 8 MHz xtal: FMPLL[0]_CR=0x02400100 56xxPBS • 40 MHz xtal: FMPLL[0]_CR=0x12400100 only) (MPC56xxP & 8 MHz xtal requires changing default CMU_CSR value. See PLL example.) Configure RUN0 Mode: • I/O Output power-down: no safe gating (default) • Main Voltage regulator is on (default) • Data, code flash in normal mode (default) • • • • PLL0 is switched on Crystal oscillator is switched on 16 MHz IRC is switched on (default) Select PLL0 (system pll) as sysclk e_b SwIrq4ISR e_b Pit1ISR DRUN=1, RUN0 = 1 ME_ME = 0x0000 001D — 8 MHz Crystal: CGM_ FMPLL[0]_CR =0x02400100 — ME_ RUN0_MC = 0x001F 0070 — ME_RUN_PC11 = 0000 0010 — ME_PCTL92 = 0x01 - ME_MCTL =0x4000 5AF0, =0x4000 A50F wait ME_GS [S_TRANS] = 0 verify 4 = ME_GS [CURRENTMODE] (See Section 10, “PLL: Initializing System Clock (MPC551x, MPC55xx)”) - — See PLL Initialization example MVRON=1 DFLAON, CFLAON= 3 PLL0ON=1 XOSC0ON=1 16MHz_IRCON=1 SYSCLK=0x4 MPC56xxB/S: • Peri. Config. 01: run in DRUN mode only RUN0=1 Assign peripheral configuration to peripherals: PIT, RTI: select ME_RUN_PC1 RUN_CFG = 0 • Verify current mode is RUN0 — PDO=0 Initiate software mode transition to RUN0 mode • Mode & key, then mode & inverted key TARGET_MODE = • Wait for transition to complete RUN0 S_TRANS CURRENTMODE init Sysclk Initialize sysclk to 64 MHz, running from PLL disable Watchdog MPC56xx BPS MPC555x Disable watchdog by writing keys to Status Register, then clearing WEN (MPC56xxBPS only) Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 69 Table 30. Initialization: INTC in Hardware Vector Mode, VLE Instructions (continued) Pseudo Code Step Relevant Bit Fields MPC551x init Irq Vectors Load the common interrupt vector prefix to spr IVPR. Value is defined in the link file. init INTC Initialize INTC: • Configure for Hardware vector mode (MPC551x note: only proc’r 0, e200z1, is used here) init PIT MPC551x: Init system clock divider for PIT, RTI to LPCLKDIV1 = 0 divide by 1 (also applies to other Group 1 (default) peripherals of eSCI A, IIC) Global controls: • Enable module • MPC555x, 56xxBPS: Freeze in debug mode spr IVPR =__IVPR_VALUE HVEN_PRC0(551x) INTC_MCR , [HVEN_PRC0] HVEN (555x) = 0 =1 MDIS = 0 FRZ = 1 Load a start count value for 64 MHz sysclk (PITs count down from value at sysclk rate) • PIT 1 Timeout = 64 M / (64 M sysclk/sec) = 1 ms Enable PIT 1 timer to count (counts down) MPC551x: Assign PIT 1 flag for IRQ instead of DMA Configure PIT1 interrupts: • MPC551x: Select processor 0 (e200z1) • Select priority 1 PRC_SEL= 0 (z1) PRI = 1 init SwIrq4 Software Interrupt 4: • Raise priority to 2 • MPC551x: Select processor(s) to interrupt PRI = 2 PRC_SEL=0 (z1) enable ExtIrq Enable recognition of requests to INTC by lowering the INTC’s current priority to 0 from default of 15 waitloop INTC_MCR [HVEN] =1 SIU_SYSCLK [LPCLKDIV1] = 0 PIT_CTRL = 0x0000 0000 PIT_PITMCR = 0x0000 0001 PIT_TVAL1 = PIT_TIMER 64,000 1_LDVAL1 = 64,000 PEN1(551x) = 1, TEN (555x, 56xxPBS)=1 TIE1 or TIE = 1 ISEL = 1 Enable PIT 1 to request IRQ MPC56xx BPS MPC555x PRI = 0 PIT_PITEN = 0x0000 0002 PIT_TIMER PIT_INTEN = 1_TCTRL = 0x0000 0002 0x0000000 PIT_INTSEL = 3 0x0000 0002 INTC_PSR [149] = 0x01 INTC_PSR [302] = 0x01 PIT_CH1_ LDVAL = 64,000 PIT_CH1_ TCTRL = 0x0000 0003 INTC_PSR [60] = 0x01 INTC_PSR[4] = 0x02 INTC_CPR _PRC0[PRI]= 0 INTC_CPR [PRI]= 0 Enable external interrupts: set MSR[EE] = 1 wrteei 1 wait forever while 1 Qorivva Simple Cookbook, Rev. 4 70 Freescale Semiconductor 7.2.2 Interrupt Handlers Table 31. Stack Frame for INTC Interrupt Handler (same as for Software Vector Mode example) Stack Frame Area 32 bit GPRs Register Location in Stack r12 sp + 0x4C r11 sp + 0x48 r10 sp + 0x44 r9 sp + 0x40 r8 sp + 0x3C r7 sp + 0x38 r6 sp + 0x34 r5 sp + 0x30 r4 sp + 0x2C r3 sp + 0x28 r0 sp + 0x24 CR CR sp + 0x20 locals and padding XER sp + 0x1C CTR sp + 0x18 LR sp + 0x14 SRR1 sp + 0x10 SRR0 sp + 0x0C padding sp + 0x08 LR area LR placeholder sp + 0x04 back chain r1 sp Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 71 Table 32. Pit 1 Interrupt Handler (INTC in Hardware Vector Mode, VLE Instructions) Relevant Bit Fields Step prolog Pseudo Code MPC551x Create stack frame MPC555x e_stwu sp, -0x50 (sp) Save SRR0:1 because nested interrupts will be allowed Re-enable external interrupts by setting MSR[EE] store SRR0:1, r3 registers to stack frame EE = 1 Save other appropriate registers for C ISR wrteei 1 store other registers to stack frame Branch to ISR, saving return address e_bl Pit1ISR Pit1ISR Increment Pit1Ctr Pit1Ctr ++ If Pit1Ctr is even, invoke Software Interrupt 4 if ((Pit1Ctr&1) ==0), INTC_SSCIR[4] = 2 SET=1 Clear Pit1Ctr interrupt flag by writing 1 to it epilog MPC56xxBPS TIF1 or TIF=1 PIT_PITFLG[TIF1] PIT_TFLG1[TIF PIT_CH1_TFLG[TIF =1 ] ] =1 =1 Restore registers from stack frame except SRR0:1 and two working registers load most registers from stack frame Ensure interrupt flag in ISR has completed clearing before writing to INTC_EOIR mbar Disable external interrupts by clearing MSR[EE] Restore former INTC’s Current Priority Restore SRR0:1 and working registers EE = 0 wrteei 0 write 0 to INTC_EOIR_PRC 0 write 0 to INTC_EOIR restore SRR0:1, working registers from stack frame Restore stack frame space e_add16i sp, sp, 0x50 Return se_rfi Qorivva Simple Cookbook, Rev. 4 72 Freescale Semiconductor Table 33. Software 4 Interrupt Handler (INTC in Hardware Vector Mode, VLE Instructions) Relevant Bit Fields Step prolog Pseudo Code MPC551x Create stack frame MPC555x e_stwu sp, -0x50 (sp) Save SRR0:1 because nested interrupts will be allowed Re-enable external interrupts by setting MSR[EE] store SRR0:1, r3 registers to stack frame EE = 1 wrteei 1 Save other appropriate registers for C ISR store other registers to stack frame Branch to ISR, saving return address bl SWIrq4ISR SWIrq4 ISR Increment SWirq4Ctr SWirq4Ctr++ epilog Restore registers from stack frame except SRR0:1 and two working registers Clear Software IRQ 4 interrupt flag by writing 1 to it CLR = 1 INTC_SSCIR4 = 1 load most registers from stack frame Ensure interrupt flag in ISR has completed clearing before writing to INTC_EOIR Disable external interrupts by clearing MSR[EE] Restore former INTC’s Current Priority Restore SRR0:1 and working registers mbar EE = 0 wrteei 0 write 0 to INTC_EOIR_PRC0 write 0 to INTC_EOIR restore SRR0:1, working registers from stack frame Restore stack frame space e_add16i sp, sp, 0x50 Return se_rfi Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 73 7.3 7.3.1 7.3.1.1 /* /* /* /* Code main.c files mainc.c (MPC551x shown with 8 MHz crystal) main.c - Hardware vector mode program using C isr */ Feb 12 2009 SM - Based on AN2865 INTC Hardware Vector Mode example*/ Aug 12 2009 SM - Changed PLL initial ERFD value, added 12MHz crystal numbers */ Copyright Freescale Semiconductor, In.c 2009 All rights reserved. */ #include "mpc5510.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ extern uint32_t __IVPR_VALUE; /* Interrupt Vector Prefix value from link file*/ extern const vuint32_t IntcIsrVectorTable[]; uint32_t Pit1Ctr = 0; /* Counter for PIT 1 interrupts */ uint32_t SWirq4Ctr = 0;/* Counter for software interrupt 4 */ void initSysclk(void) { /* Initialize PLL and sysclk to 64 MHz */ /* Use 2 of the next 4 lines: */ FMPLL.ESYNCR2.R = 0x00000007; /* 8MHz xtal: ERFD to initial value of 7 */ FMPLL.ESYNCR1.R = 0xF0000020; /* 8MHz xtal: CLKCFG=PLL, EPREDIV=0, EMFD=0x20*/ /*FMPLL.ESYNCR2.R = 0x00000005; */ /* 12MHz xtal: ERFD to initial value of 5 */ /*FMPLL.ESYNCR1.R = 0xF0020030; */ /* 12MHz xtal: CLKCFG=PLL, EPREDIV=2, EMFD=0x30*/ CRP.CLKSRC.B.XOSCEN = 1; /* Enable external oscillator */ while (FMPLL.SYNSR.B.LOCK != 1) {}; /* Wait for PLL to LOCK */ /* Use 1 of the next 2 lines: */ FMPLL.ESYNCR2.R = 0x00000005; /* 8MHz xtal: ERFD change for 64 MHz sysclk */ /*FMPLL.ESYNCR2.R = 0x00000003; */ /* 12MHz xtal: ERFD change for 64 MHz sysclk */ SIU.SYSCLK.B.SYSCLKSEL = 2; /* Select PLL for sysclk */ } asm void initIrqVectors(void) { lis r3, __IVPR_VALUE@h /* IVPR value is passed from link file */ ori r3, r3, __IVPR_VALUE@l mtivpr r3 } void initINTC(void) { INTC.MCR.B.HVEN_PRC0 = 1; /* MPC551x Proc'r 0: initialize for HW vector mode*/ } void initPIT(void) { SIU.SYSCLK.B.LPCLKDIV1 = 0; PIT.PITCTRL.R = 0; PIT.TLVAL[1].R = 64000; PIT.PITINTEN.R = 0x00000002; PIT.PITINTSEL.R = 0x00000006; PIT.PITEN.B.PEN1 = 1; INTC.PSR[149].R = 0x01; } /* /* /* /* /* /* /* Divide sysclk by 1 for Group 1 modules */ Ensure PIT is not disbaled */ Timeout= 64K sysclks x 1sec/64M sysclks= 1 ms */ Enable PIT 1 flag to request INTC or DMA */ Assign PIT 1 flag to select IRQ, not DMA req. */ Start PIT counting */ PIT 1 interrupt selects proc'r 0 & priority 1 */ void initSwIrq4(void) { INTC.PSR[4].R = 2;/* Software interrupt 4 IRQ priority = 2 */ } void enableIrq(void) { INTC.CPR_PRC0.B.PRI = 0; asm(" wrteei 1"); } /* MPC551x Proc'r 0: Lower INTC's current priority */ /* Enable external interrupts */ Qorivva Simple Cookbook, Rev. 4 74 Freescale Semiconductor void main (void) { vuint32_t i = 0; initSysclk(); initIrqVectors(); initINTC(); initPIT(); initSwIrq4(); enableIrq(); } /* Dummy idle counter */ /* /* /* /* /* /* Set sysclk to 64 MHz */ Initialize exceptions: only need to load IVPR */ Initialize INTC for software vector mode */ Initialize PIT1 for 1kHz IRQ, priority 2 */ Initialize software interrupt 4 */ Ensure INTC current prority=0 & enable IRQ */ while (1) { i++; } void Pit1ISR(void) { Pit1Ctr++; if ((Pit1Ctr & 1)==0) { INTC.SSCIR[4].R = 2; } PIT.PITFLG.B.TIF1 = 1; } void SwIrq4ISR(void) { SWirq4Ctr++; INTC.SSCIR[4].R = 1; } /* Increment interrupt counter */ /* If PIT1Ctr is even*/ /* then invoke software interrupt 4 */ /* Clear PIT 1 flag by writing 1 */ /* Increment interrupt counter */ /* Clear channel's flag */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 75 7.3.1.2 main.c (MPC555x MPC563xM shown with 8 MHz crystal) /* main.c - Hardware vector mode program using C isr */ /* Feb 03 2009 SM - Based on AN2865 INTC Software Vector Mode example*/ /* Copyright Freescale Semiconductor, In.c 2009 All rights reserved. */ #include "mpc563m.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ extern uint32_t __IVPR_VALUE; /* Interrupt Vector Prefix value from link file*/ uint32_t Pit1Ctr = 0; /* Counter for PIT 1 interrupts */ uint32_t SWirq4Ctr = 0;/* Counter for software interrupt 4 */ void initSysclk(void) { /* Intialize sysclk to 64MHz for 8 MHz crystal*/ FMPLL.ESYNCR1.B.CLKCFG = 0X7; /* Change clk to PLL normal mode from crystal */ FMPLL.SYNCR.R = 0x16080000; /* Initial values: PREDIV=1, MFD=12, RFD=1 */ while (FMPLL.SYNSR.B.LOCK != 1) {}; /* Wait for FMPLL to LOCK */ FMPLL.SYNCR.R = 0x16000000; /* Final value for 64 MHz: RFD=0 */ } asm void initIrqVectors(void) { lisr3, __IVPR_VALUE@h /* IVPR value is passed from link file */ ori r3, r3, __IVPR_VALUE@l /* Note: IVPR lower bits are unused in MPC555x */ mtivprr3 } void initINTC(void) { INTC.MCR.B.HVEN = 1; } /* Single core: Initialize for HW vector mosde */ void initPIT(void) { PIT.MCR.R = 0x00000001; /* Enable PIT module & freeze count during debug */ PIT.TIMER[1].LDVAL.R = 64000; /* Timeout= 64K sysclks x 1sec/64M sysclks= 1 ms */ PIT.TIMER[1].TCTRL.R = 0x00000003; /* Start timer counting & freeze during debug */ INTC.PSR[302].R = 0x1; /* PIT 1 interrupt has priority 1 */ } void initSwIrq4(void) { INTC.PSR[4].R = 2; /* Software interrupt 4 IRQ priority = 2 */ } void enableIrq(void) { INTC.CPR.B.PRI = 0; asm(" wrteei 1"); } /* Lower INTC's current priority */ /* Enable external interrupts */ void main (void) { vuint32_t i = 0;/* Dummy idle counter */ } initSysclk(); initIrqVectors(); initINTC(); initPIT(); initSwIrq4(); enableIrq(); while (1) { i++; } /* /* /* /* /* /* Set sysclk to 64 MHz */ Initialize exceptions: only need to load IVPR */ Initialize INTC for software vector mode */ Initialize PIT1 for 1kHz IRQ, priority 2 */ Initialize software interrupt 4 */ Ensure INTC current prority=0 & enable IRQ */ void Pit1ISR(void) { Pit1Ctr++; if ((Pit1Ctr & 1)==0) { INTC.SSCIR[4].R = 2; } PIT.TIMER[1].TFLG.B.TIF = 1; } void SwIrq4ISR(void) { SWirq4Ctr++; INTC.SSCIR[4].R = 1; } /* Increment interrupt counter */ /* If PIT1Ctr is even*/ /* then nvoke software interrupt 4 */ /* Clear PIT 1 flag by writing 1 */ /* Increment interrupt counter */ /* Clear channel's flag */ Qorivva Simple Cookbook, Rev. 4 76 Freescale Semiconductor 7.3.1.3 /* /* /* /* /* /* main.c (MPC56xB/P/S - MPC56xxS shown with 8 MHz crystal) main.c - Hardware vector mode program using C isr */ Feb 12, 2009 S.Mihalik Initial version based on previous AN2865 example */ May 22 2009 S. Mihalik- Simplifed by removing unneeded sysclk, PCTL code */ Jul 03 2009 S Mihalik - Simplified code */ Mar 15 2010 SM - modified initModesAndClks, updated header */ Copyright Freescale Semiconductor, Inc. 2009, 2010. All rights reserved. */ #include "56xxS_0204.h" /* Use proper include file */ extern uint32_t __IVPR_VALUE; /* Interrupt Vector Prefix value from link file*/ uint32_t Pit1Ctr = 0; /* Counter for PIT 1 interrupts */ uint32_t SWirq4Ctr = 0;/* Counter for software interrupt 4 */ void initModesAndClock(void) { ME.MER.R = 0x0000001D; } /* Enable DRUN, RUN0, SAFE, RESET modes */ /* Initialize PLL before turning it on: */ CGM.FMPLL[0].CR.R = 0x02400100; /* 8 MHz xtal: Set PLL0 to 64 MHz */ ME.RUN[0].R = 0x001F0074; /* RUN0 cfg: 16MHzIRCON,OSC0ON,PLL0ON,syclk=PLL */ ME.RUNPC[1].R = 0x00000010; /* Peri. Cfg. 1 settings: only run in RUN0 mode */ ME.PCTL[92].R = 0x01; /* PIT, RTI: select ME_RUN_PC[1] */ /* Mode Transition to enter RUN0 mode: */ ME.MCTL.R = 0x40005AF0; /* Enter RUN0 Mode & Key */ ME.MCTL.R = 0x4000A50F; /* Enter RUN0 Mode & Inverted Key */ while (ME.GS.B.S_MTRANS) {} /* Wait for mode transition to complete */ /* Note: could wait here using timer and/or I_TC IRQ */ while(ME.GS.B.S_CURRENTMODE != 4) {} /* Verify RUN0 is the current mode */ void disableWatchdog(void) { SWT.SR.R = 0x0000c520; /* Write keys to clear soft lock bit */ SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; /* Clear watchdog enable (WEN) */ } asm void initIrqVectors(void) { lis r3, __IVPR_VALUE@h /* IVPR value is passed from link file */ ori r3, r3, __IVPR_VALUE@l mtivpr r3 } void initINTC(void) { INTC.MCR.B.HVEN = 1; } /* Single core: initialize for HW vector mode */ void initPIT(void) { PIT.MCR.R = 0x00000001; /* Enable PIT and configure to stop in debug mode */ PIT.CH[1].LDVAL.R = 64000; /* Timeout= 64000 sysclks x 1sec/64M sysclks = 1 ms*/ PIT.CH[1].TCTRL.R = 0x000000003; /* Enable PIT1 interrupt & start PIT counting */ INTC.PSR[60].R = 0x01; /* PIT 1 interrupt vector with priority 1 */ } void initSwIrq4(void) { INTC.PSR[4].R = 2; } /* Software interrupt 4 IRQ priority = 2 */ void enableIrq(void) { INTC.CPR.B.PRI = 0; asm(" wrteei 1"); } /* Single Core: Lower INTC's current priority */ /* Enable external interrupts */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 77 void main (void) { vuint32_t i = 0; initModesAndClock(); disableWatchdog(); initIrqVectors(); initINTC(); initPIT(); initSwIrq4(); enableIrq(); } /* Dummy idle counter */ /* /* /* /* /* /* /* MPC56xxP/B/S: Initialize mode entries, set sysclk=64 MHz*/ Disable watchdog */ Initialize exceptions: only need to load IVPR */ Initialize INTC for software vector mode */ Initialize PIT1 for 1kHz IRQ, priority 2 */ Initialize software interrupt 4 */ Ensure INTC current prority=0 & enable IRQ */ while (1) { i++; } void Pit1ISR(void) { Pit1Ctr++; if ((Pit1Ctr & 1)==0) { INTC.SSCIR[4].R = 2; } PIT.CH[1].TFLG.B.TIF = 1; } void SwIrq4ISR(void) { SWirq4Ctr++; INTC.SSCIR[4].R = 1; } /* Increment interrupt counter */ /* If PIT1Ctr is even*/ /* then nvoke software interrupt 4 */ /* MPC56xxP/B/S: CLear PIT 1 flag by writing 1 */ /* Increment interrupt counter */ /* Clear channel's flag */ Qorivva Simple Cookbook, Rev. 4 78 Freescale Semiconductor 7.3.2 # # # # # # # # # # handlers_vle.s file handlers_vle.s - INTC hardware vector mode example using VLE instructions Description: Creates prolog, epilog for C ISR and enables nested interrupts Rev 1.0: April 23, 2004, S Mihalik, Rev 1.1 Aug 2, 2004 SM - delayed writing to EOIR until after disabling EE in epilog Rev 1.2 Sept 8 2004 SM - optimized & corrected r3,r4 restore sequence from rev 1.1 Rev 1.2 Sept 21 2004 SM - optimized by minimizing time interrupts are disabled Rev 1.3 Jul 2 2007 SM - Changes for MPC551x and mapped to .ivor_handlers section Rev 1.4 Jan 22 2009 SM - Modified for VLE instructions, CodeWarrior 2.4 alpha Rev 1.5 Mar 09 2010 SM - Removed unneeded Epilogue instruction: se_mtlr r3 Copyright Freescale Semiconductor, Inc. 2009, 2010. All rights reserved # STACK FRAME DESIGN: Depth: 20 words (0xA0, or 80 bytes) # ************* ______________ # 0x4C * GPR12 * ^ # 0x48 * GPR11 * | # 0x44 * GPR10 * | # 0x40 * GPR9 * | # 0x3C * GPR8 * | # 0x38 * GPR7 * GPRs (32 bit) # 0x34 * GPR6 * | # 0x30 * GPR5 * | # 0x2C * GPR4 * | # 0x28 * GPR3 * | # 0x24 * GPR0 * ___v__________ # 0x20 * CR * __CR__________ # 0x1C * XER * ^ # 0x18 * CTR * | # 0x14 * LR * locals & padding for 16 B alignment # 0x10 * SRR1 * | # 0x0C * SRR0 * | # 0x08 * padding * ___v__________ # 0x04 * resvd- LR * Reserved for calling function # 0x00 * SP * Backchain (same as gpr1 in GPRs) # ************* .section .ivor_handlers,text_vle .equINTC_EOIR_PRC0, 0xfff48018 # Dual Core: Proc 0 End Of Interrupt Reg. addr. .equINTC_EOIR, 0xfff48018 # Single Core: End Of Interrupt Reg. addr. .extern Pit1ISR .extern SwIrq4ISR .globl Pit1Handler .globl SwIrq4Handler Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 79 Pit1Handler: e_stwu se_stw r1, -0x50 (r1) r3, 0x28 (r1) mfsrr0 se_stw mfsrr1 se_stw r3 r3, r3 r3, wrteei 1 e_stw r12, e_stw r11, e_stw r10, e_stw r9, e_stw r8, se_stw r7, se_stw r6, se_stw r5, se_stw r4, se_stw r0, mfcr r3 se_stw r3, mfxer r3 se_stw r3, se_mfctr r3 se_stw r3, se_mflr r4 se_stw r4, e_bl 0x0C (r1) 0x10 (r1) 0x4C 0x48 0x44 0x40 0x3C 0x38 0x34 0x30 0x2C 0x24 wrteei # e_stw e_stw 0x1C (r1) 0x18 (r1) 0x14 (r1) Pit1ISR # Set MSR[EE]=1 # Store rest of gprs # Store CR # Store XER # Store CTR # Store LR # Branch to ISR, but return here (r1) # EPILOGUE # Restore LR (r1) # Restore CTR (r1) # Restore XER (r1) # Restore CR (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) # Restore other gprs except working registers # # r3, INTC_EOIR_PRC0@ha # r3, INTC_EOIR@ha # r4, 0 0 Ensure store to clear interrupt flag bit completed Use 1 of the following 2 lines: Dual Core: Load upper half proc 0 EIOR addr to r3 Single Core: Load upper half of EIOR address to r3 # # r4, INTC_EOIR_PRC0@l(r3) # r4, INTC_EOIR@l(r3) # se_lwz r3, mtsrr0 r3 se_lwz r3, mtsrr1 r3 se_lwz r4, se_lwz r3, e_add16i r1, se_rfi (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) 0x20 (r1) se_lwz r3, 0x14 se_mtlr r3 se_lwz r3, 0x18 se_mtctr r3 se_lwz r3, 0x1C mtxer r3 se_lwz r3, 0x20 mtcrf 0xff, r3 se_lwz r0, 0x24 se_lwz r5, 0x30 se_lwz r6, 0x34 se_lwz r7, 0x38 e_lwz r8, 0x3C e_lwz r9, 0x40 e_lwz r10, 0x44 e_lwz r11, 0x48 e_lwz r12, 0x4C mbar 0 # e_lis e_lis se_li # PROLOGUE # Create stack frame and store back chain # Store a working register # Note: use se_ form for r0-7, r24-31 with positive offset # Store SRR0:1 (must be done before enabling EE) 0x0C (r1) Disable interrupts for rest of handler Use 1 or 2 of the next appropriate lines: Dual Core - Write 0 to proc'r 0 INTC_EIOR Single Core - Write 0 to proc'r 0 INTC_EIOR # Restore SRR0 0x10 (r1) # Restore SRR1 0x2C (r1) 0x28 (r1) r1, 0x50 # Restore working registers # Delete stack frame # End of Interrupt Qorivva Simple Cookbook, Rev. 4 80 Freescale Semiconductor SwIrq4Handler: e_stwu se_stw r1, -0x50 (r1) r3, 0x28 (r1) mfsrr0 se_stw mfsrr1 se_stw r3 r3, r3 r3, wrteei 1 e_stw r12, e_stw r11, e_stw r10, e_stw r9, e_stw r8, se_stw r7, se_stw r6, se_stw r5, se_stw r4, se_stw r0, mfcr r3 se_stw r3, mfxer r3 se_stw r3, se_mfctr r3 se_stw r3, se_mflr r4 se_stw r4, e_bl 0x0C (r1) 0x10 (r1) 0x4C 0x48 0x44 0x40 0x3C 0x38 0x34 0x30 0x2C 0x24 wrteei # e_stw e_stw (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) 0x20 (r1) 0x1C (r1) 0x18 (r1) 0x14 (r1) SwIrq4ISR se_lwz r3, 0x14 se_mtlr r3 se_lwz r3, 0x18 se_mtctr r3 se_lwz r3, 0x1C mtxer r3 se_lwz r3, 0x20 mtcrf 0xff, r3 se_lwz r0, 0x24 se_lwz r5, 0x30 se_lwz r6, 0x34 se_lwz r7, 0x38 e_lwz r8, 0x3C e_lwz r9, 0x40 e_lwz r10, 0x44 e_lwz r11, 0x48 e_lwz r12, 0x4C mbar 0 # e_lis e_lis se_li # PROLOGUE # Create stack frame and store back chain # Store a working register # Note: use se_ form for r0-7, r24-31 with positive offset # Store SRR0:1 (must be done before enabling EE) # Set MSR[EE]=1 # Store rest of gprs # Store CR # Store XER # Store CTR # Store LR # Branch to ISR, but return here (r1) # EPILOGUE # Restore LR (r1) # Restore CTR (r1) # Restore XER (r1) # Restore CR (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) (r1) # Restore other gprs except working registers # # r3, INTC_EOIR_PRC0@ha # r3, INTC_EOIR@ha # r4, 0 0 Ensure store to clear interrupt flag bit completed Use 1 of the following 2 lines: Dual Core: Load upper half proc 0 EIOR addr to r3 Single Core: Load upper half of EIOR address to r3 # # r4, INTC_EOIR_PRC0@l(r3) # r4, INTC_EOIR@l(r3) # se_lwz r3, mtsrr0 r3 se_lwz r3, mtsrr1 r3 se_lwz r4, se_lwz r3, e_add16i r1, 0x0C (r1) Disable interrupts for rest of handler Use 1 or 2 of the next appropriate lines: Dual Core - Write 0 to proc'r 0 INTC_EIOR Single Core - Write 0 to proc'r 0 INTC_EIOR # Restore SRR0 0x10 (r1) # Restore SRR1 0x2C (r1) 0x28 (r1) r1, 0x50 # Restore working registers se_rfi # Delete stack frame # End of Interrupt Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 81 7.3.3 # # # # intc_hw_branch_table.s file (MPC56xxB/P/S vectors shown) Rev 1.0 Jul 2, 2007 S Mihalik Rev 1.1 Aug 30 1007 SM - Made SwIrq4Handler, emiosCh0Handler .extern Rev 2.0 Jan 22 2009 SM - Modified for VLE and MPC56xxB/P/S INTC vector numbers Copyright Freescale Semiconductor, Inc. 2007. All rights reserved .section .intc_hw_branch_table,text_vle .extern SwIrq4Handler .extern Pit1Handler .equ ALIGN_OFFSET, 4 # MPC551x,MPC56xxB/P/S: 4 byte branch alignments (Diab/GHS use 2; CW 4) #.equ ALIGN_OFFSET, 4 # MPC555x: 16 byte branch alignments (Diab/GHS use 4; CW 16) IntcHandlerBranchTable: # Only 100 example vectors are implemented here # MPC555x: This table must have 64 KB alignment # MPC551x, MPC56xxB/P/S: Requires 2 KB alignment after 4KB boundary hw_vect0: hw_vect1: hw_vect2: hw_vect3: hw_vect4: hw_vect5: .align ALIGN_OFFSET e_b hw_vect0 .align ALIGN_OFFSET e_b hw_vect1 .align ALIGN_OFFSET e_b hw_vect2 .align ALIGN_OFFSET e_b hw_vect3 .align ALIGN_OFFSET e_b SwIrq4Handler .align ALIGN_OFFSET e_b hw_vect5 #INTC HW vector 0 #INTC HW vector 1 #INTC HW vector 2 #INTC HW vector 3 # SW IRQ 4 #INTC HW vector 5 ... etc. for contiguous vectors ........................................................ hw_vect57: hw_vect58: hw_vect59: hw_vect61: hw_vect62: hw_vect63: hw_vect64: .align ALIGN_OFFSET e_b hw_vect57 #INTC HW vector 57 .align ALIGN_OFFSET e_b hw_vect58 #INTC HW vector 58 .align ALIGN_OFFSET e_b hw_vect59 #INTC HW vector 59 .align ALIGN_OFFSET e_b Pit1Handler #INTC HW vector 60 .align ALIGN_OFFSET e_b hw_vect61 #INTC HW vector 61 .align ALIGN_OFFSET e_b hw_vect62 #INTC HW vector 62 .align ALIGN_OFFSET e_b hw_vect63 #INTC HW vector 63 .align ALIGN_OFFSET e_b hw_vect64 #INTC HW vector 64 ... etc. for other vectors ........................................................ Qorivva Simple Cookbook, Rev. 4 82 Freescale Semiconductor 7.3.4 # # # # # # ivor_branch_table.s file (MPC551x, MPC56xxB/P/S only) ivor_branch_table.s - for use with MPC551x, MPC56xxP, MPC56xxB, MPC56xxS only Description: Branch table for 16 ore interrupts Copyright Freescale 2007, 2009. All Rights Reserved Rev 1.0 Jul 6 2007 S Mihalik - Initial version Rev 1.1 Aug 30 2007 SM - Made IVOR4Handler extern (for SW vector mode) Rev 1.2 Sep 9 2008 SM - Converted assembly to VLE syntax .extern IVOR4Handler .section .ivor_branch_table,text_vle .equ SIXTEEN_BYTES, 16 IVOR0trap: IVOR1trap: IVOR2trap: IVOR3trap: IVOR3trap: IVOR5trap: IVOR6trap: IVOR7trap: IVOR8trap: IVOR9trap: IVOR10trap: IVOR11trap: IVOR12trap: IVOR13trap: IVOR14trap: IVOR15trap: # 16 byte alignment required for table entries # Diab compiler uses value of 4 (2**4=16) # CodeWarrior, GHS, Cygnus use 16 .align SIXTEEN_BYTES e_b IVOR0trap # IVOR 0 interrupt handler .align SIXTEEN_BYTES e_b IVOR1trap # IVOR 1 interrupt handler .align SIXTEEN_BYTES e_b IVOR2trap # IVOR 2 interrupt handler .align SIXTEEN_BYTES e_b IVOR3trap # IVOR 3 interrupt handler .align SIXTEEN_BYTES e_b IVOR4trap # IVOR 4 interrupt handler .align SIXTEEN_BYTES e_b IVOR5trap # IVOR 5 interrupt handler .align SIXTEEN_BYTES e_b IVOR6trap # IVOR 6 interrupt handler .align SIXTEEN_BYTES e_b IVOR7trap # IVOR 7 interrupt handler .align SIXTEEN_BYTES e_b IVOR8trap # IVOR 8 interrupt handler .align SIXTEEN_BYTES e_b IVOR9trap # IVOR 9 interrupt handler .align SIXTEEN_BYTES e_b IVOR10trap # IVOR 10 interrupt handler .align SIXTEEN_BYTES e_b IVOR11trap # IVOR 11 interrupt handler .align SIXTEEN_BYTES e_b IVOR12trap # IVOR 12 interrupt handler .align SIXTEEN_BYTES e_b IVOR13trap # IVOR 13 interrupt handler .align SIXTEEN_BYTES e_b IVOR14trap # IVOR 14 interrupt handler .align SIXTEEN_BYTES e_b IVOR15trap # IVOR15 interrupt handler Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 83 8 MMU: Create TLB Entry 8.1 Description Task: Initialize translation lookaside buffer (TLB) entry number six to define a memory page defined below. Table 34. TLB Entry 6 Summary Page Attribute Value Relevant BIt Field Starting Effective Address 0x4004 0000 EPN Size 4 KB SIZE Translation No translation RPN = EPN Access allowed Data, both R + W for all UR, SR, UW, SW = 1 Access not allowed Executable UX, SX = 0 Write Mode Policy1 Not write through W=0 Cache Inhibit Not inhibited I=0 Coherency, Guarded, Endianness Default M, G, E = 0 Global Process ID Default TID = 0 Translation Space Default TS = 0 1 To allow MMU entries to specify the write mode policy, the cache write mode, L1CSR0[CWM], must be set to 1. This program shows how MMU entries are created. However, there is not any memory or I/O on current MPC5500 devices for the effective address range of this entry, so nothing can be accessed yet. The cache program will later use this entry to allocate part of cache as SRAM. Accessing the MMU’s TLB entries is done using sprs MAS0:3, as shown below. MPC5500 gpr gpr mtspr mfspr spr’s MAS1:3 spr’s MAS1:3 tlbwe (entry number defined in MAS0[ESEL]) MMU Entry 00 Entry 01 ••• Entry n tlbre* TLB Table V TS TID EPN RPN SIZE Permissions Memory/Cache Attributes User Attributes IPROT V TS TID EPN RPN SIZE Permissions Memory/Cache Attributes User Attributes IPROT ••• V TS TID EPN RPN SIZE Permissions Memory/Cache Attributes User Attributes IPROT Figure 18. MMU TLB Entry Example Qorivva Simple Cookbook, Rev. 4 84 Freescale Semiconductor Exercise: If your debugger supports viewing the MMU TLB table, view the table entries before and after running the code1. Create a new MMU entry seven that starts at address 0x4009 0000 and has a size of 16 KB. 8.2 Design Below is a version of figure 6-6 of the e200z6 PowerPC™ Core Reference Manual Rev. 0 for reference. 00 MAS0 01 0 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 TBSEL 0 ESEL MAS1 VALID IPROT TID MAS2 EPN MAS3 RPN 0 0 TS NV TSIZ 0 0 0 VLE W I M G E U0 U1 U2 U3 UX SX UW SW UR SR Figure 19. MMU Assist Registers 0–3 Summary Table 35. Initialization: MMU TLB Entry 6 Step Relevant Bit Fields Initialize MMU Select TLB entry number and define MMU R/W & TLB entry 6 Replacement Control: • Select TLB entry 6 • Select TLB 1 (required) • Null replacement spr MAS0 = 0x1006 0000 MAS0[ESEL] = 6 MAS0[TBLSEL] = 1 MAS0[NVCAM] = 0 Define Descriptor Context and Configuration Control: • Set page as valid • No invalidation protection • Use global process ID • Use default translation space • Size = 4KB MAS1[VALID] = 1 MAS1[IPROT] = 0 MAS1[TID] = 0 MAS1[TS] = 0 MAS1[TSIZE] – 0x1 Define EPN and Page Attributes: • EPN = 0x40 0400 for address 0x4004 0000 • WIMAGE = all 0’s; use cache inhibit, no write-thru MAS2[EPN] = 0x40 0400 MAS2[W, I, M, G, E] = 0 Define RPN and Access Control (here as RW data) • RPN = 0x40 0400 for address 0x4004 0000 • Set user bits to 0 (unused) • Disable both user & supervisor execution • Enable both user and supervisor data R and W MAS3[RPN] = 0X40 0400 MAS3[U0:3] = 0 MAS3[UX, SX] = 0 MAS3[UR, SR, UW, SW] = 1 Write TLB entry fields in MAS1:3 to the entry per MAS0[ESEL] 1 Pseudo Code spr MAS1 = 0x8000 0100 spr MAS2 = 0x4004 0000 spr MAS3 = 0x4004 000F tlbwe1 Per Book E, a context synchronizing instruction (CSI), such as an isync, would be needed before the tlbwe instruction if there were uncompleted preceding storage accesses. Similarly a CSI, such as isync or mbar, would be needed after the tlbwe to ensure that any immediate subsequent storage accesses use the updated TLB entry values. 1. To display the current MMU table in a debugger: Green Hills — type “target tlbr 0..7” from the command prompt for TLB entries 0 to 7; Lauterbach — select “MMU TLB1 Table” from the MPC5554 menu; PEMicro — hit the MMU button or type “showmmu.” Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 85 8.3 /* /* /* /* /* /* Code main.c - MMU - create TLB entry example */ Description: Creates a new entry to the TLB table in the MMU */ Rev 1.0 Sept 28, 2003 S.Mihalik, Copyright Freescale, 2004. All Rights Reserved */ Notes: */ 1. MMU not initialized; must be done by debug scripts or BAM */ 2. L2SRAM not initialized; must be done by debug scripts */ #include "mpc563m.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ asm void MMU_init_TLB6(void) { lis r3, 0x1006 /* Select TLB entry #, define R/W replacment control */ mtMAS0 r3 /* Load MAS0 with 0x1006 0000 for TLB entry #6 */ lis r3, 0x8000 ori r3, r3, 0x0100 mtMAS1 r3 lis r3, 0x4004 mtMAS2 r3 lis r3, 0x4004 ori r3, r3, 0x000F mtMAS3 r3 } tlbwe void main (void) { int i = 0; } MMU_init_TLB6(); while (1) { i++; } /* Define description context and configuration control:*/ /* VALID=1, IPROT=0, TID=0, TS=0, TSIZE=1 (4KB size) */ /* Load MAS 1 with 0x8000 0100 */ /* Define EPN and page attributes: */ /* EPN = 0x4004 0000, WIMAGE = all 0's */ /* Load MAS2 with 0x4004 0000 */ /* Define RPN and access control for data R/W */ /* RPN = 0x4004 0000, U0:3=0, UX/SX=0, UR/SR/UW/SW=1 */ /* Load MAS3 with 0x4004 000F */ /* Write entry defined in MAS0 (entry 6 here) to MMU TLB */ /* Dummy idle counter */ /* Define 4KB R/W space starting at 0x4004 0000 */ /* Loop forever */ Qorivva Simple Cookbook, Rev. 4 86 Freescale Semiconductor 9 Cache: Cache as RAM 9.1 Description Task: Use 4 KB of cache as RAM. Use the memory page created in Section 8, “MMU: Create TLB Entry,” that uses memory starting at address 0x4004 0000 and that is 4 KB in size. Note that cache is not available on all MPC555x devices. It is available on those with the e200z6 core. This example shows how to turn the cache on (enable) based on information from the e200z6 PowerPCTM Core Reference Manual, Rev 0. After it is enabled, part of the cache is allocated to be used as RAM. The advantages of using cache as RAM is that it increases available memory to a program and is the fastest type of memory. Normally cache initialization is done shortly after reset, well before main and C code execution. Although this example is implemented in a C program, the cache as well as MMU entries would normally be initialized in an assembler program as part of initialization. See AN2789, MPC5500 Configuration and Initialization, for further information. Cache invalidation takes approximately 134 cycles for the 32 KB cache of the MPC5554. This design simply polls the INV bit to wait for an indication that invalidation completed, i.e., L1SR0[CINV] = 0. However, during this time it would be possible to perform some other task, or use other resources such as DMA. Exercise: Execute the code. View the MMU entry created (if your debugger supports this feature). View and modify memory created in cache1. Add instructions to lock some data or code in cache. (Hint: Point a gpr to the address of data or code to be locked, then have LockingLoop use either dcbtls or icbtls instruction for each cache line.) MPC5500 Cache Inhibit = 1 e200z6 Core Effective Address MMU Real Address CPU Core Cache Inhibit = 0 Real Address Cache Data/Instructions Data/ Instructions Memory Cache Bypassed Figure 20. Cache as RAM Example 1. Use the debugger’s memory display window with the address of the new memory created. Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 87 9.2 Design Table 36. Cache: Configure Part as RAM Relevant Bit Fields Step Enable cache If cache is already being invalidated, wait until done. Pseudo Code CINV = 0 Wait for L1SCR0[CINV] = 0 • Set cache invalidation bit CINV = 1 msync isync spr L1CSR0[CINV] = 1 Wait until cache invalidation completes. CINV = 0 Wait for L1SCR0[CINV] = 0 If cache invalidation was aborted, then attempt another cache invalidation. ABT= 1 Invalidate cache. • Use sync instructions before writing to L1CSR0 Enable cache in copyback mode, keeping other control bits at 0. • Use sync instructions before writing to L1CSR0 • Set mode to copyback • Enable cache CWM = 1 CE = 1 Initialize (See previous example, Section 8, “MMU: Create TLB Entry.”) MMU TLB entry 6 if L1SCR0[ABT] = 1, then try to invalidate cache again msync isync L1SCR0 = 0x0010 0001 spr MAS0 = 0x1006 0000 spr MAS1 = 0x8000 0100 spr MAS2 = 0x4004 0000 spr MAS3 = 0x4004 000F tlbwe1 Lock 4KB Initialize loop counter for 4 KB block size / 32 byte line size = 128 in cache Load a register with start address of new address space. r3 = 0x4004 0000 For each 32-byte cache line in the 4 KB space: • Establish line address in cache (and zero out line) • Lock line in cache Locking Loop: dcbz instruction for line’s address dcbtls instruction for line’s address 1 spr CTR = 0x80 Per Book E, a context synchronizing instruction (CSI), such as an isync, would be needed before the tlbwe instruction if there were uncompleted preceding storage accesses. Similarly a CSI, such as isync or mbar, would be needed after the tlbwe to ensure that any immediate subsequent storage accesses use the updated TLB entry values. CAUTION: Do not execute cache-based code that reconfigures the cache which could overwrite the cached code being executed! Example improper sequence is: 1. Out of reset, the flash MMU page has Cache Inhibit = 0, so when the cache is enabled code will be put in cache 2. Cache is enabled. 3. MMU TLB entry is initialized as above example. 4. Cache line addresses are established and locked into cache. Possible result: If the code just happens to be in the cache line that is being converted to be used as SRAM, then that code could be erased. Solution: Before enabling cache, change the flash MMU page from CI = 0 to CI = 1, so flash is temporarily cache inhibited. After the cache is enabled and configured, then change the flash MMU page back to CI = 0. Qorivva Simple Cookbook, Rev. 4 88 Freescale Semiconductor 9.3 Code /* main.c - Configure part of cache as SRAM */ /* Rev 0.1 Sept 27, 2004 S.Mihalik, Copyright Freescale, 2004. All Rights Reserved */ /* Notes: */ /* 1. MMU not initialized; must be done by debug scripts or BAM */ /* 2. L2SRAM not initialized; must be done by debug scripts */ #include "mpc563m.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ asm void cache_enable(void) { WaitForInvalidationComplete1: /* If CINV is already set, wait until it clears */ mfspr r3, l1csr0 /* Get current status of cache from spr L1CSR0*/ li r4, 0x0002 /* In a scratch register set bit 30 to 1, other bits 0 */ and r3, r3, r4 /* Mask out all bits except CINV*/ cmpli r3, 0x0002 /* Compare if CINV, bit 30, =1 */ beq WaitForInvalidationComplete1 InvalidateCache: msync isync mtspr l1csr0, r4 /* Before writing to L1CSR0, execute msync & isync */ /* Invalidate cache: Set L1CSR0[CINV]=1; other fields=0*/ WaitForInvalidationComplete2: mfspr r3, l1csr0 /* Get current status of cache from spr L1CSR0*/ and r3, r3, r4 /* Mask out all bits except CINV */ cmpli r3, 0x0002 /* Compare if CINV, bit 30, =1 */ beq WaitForInvalidationComplete2 } mfspr li and cmpli beq r3, l1csr0 r4, 0x0004 r3, r3, r4 r3, 0x0004 InvalidateCache /* /* /* /* /* /* Branch to error if cache invalidation was aborted */ Get current status of cache from spr L1CSR0*/ In a scratch register set bit 29 to 1, other bits 0 */ Mask out all bits except CABT*/ Compare if CABT, bit 29, = 1 */ If there was an aborted invalidation, attempt again */ lis ori msync isync mtspr r3, 0x0010 r3, r3, 0x1 /* Enable cache */ /* In a scratch register set bit 12 = 0 (used to set CWM)*/ /* Also set bit 31 =1 (used for setting CE) */ /* Before writing to L1CSR0, execute msync & isync */ l1csr0, r3 /* Enable cache with Cache others to 0 */ asm void MMU_init_TLB6(void) { lis r3, 0x1006 /* Select TLB entry #, define R/W replacment control */ mtMAS0 r3 /* Load MAS0 with 0x1006 0000 for TLB entry #6 */ /* Define description context and configuration control:*/ /* VALID=1, IPROT=0, TID=0, TS=0, TSIZE=1 (4KB size) */ lis r3, 0x8000 /* Load MAS 1 with 0x8000 0100 */ ori r3, r3, 0x0100 mtMAS1 r3 /* Define EPN and page attributes: */ /* EPN = 0x4004 0000, WIMAGE = all 0's */ lis r3, 0x4004 /* Load MAS2 with 0x4004 0000 */ mtMAS2 r3 /* Define RPN and access control for data R/W */ /* RPN = 0x4004 0000, U0:3=0, UX/SX=0, UR/SR/UW/SW=1 */ lis r3, 0x4004 /* Load MAS3 with 0x4004 000F */ ori r3, r3, 0x000F mtMAS3 r3 tlbwe /* Write entry defined in MAS0 (entry 6 here) to MMU TLB */ } asm void li mtCTR lis cache_lock_4KB(void) { r3, 0x80 /* Load r3 with loop count = 4KB/32B = 128 = 0x80 */ r3 /* Move loop count to spr CTR */ r3, 0x4004 /* Point r3 to start of desired address (r3=0x40040000)*/ LockingLoop: Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 89 } dcbz dcbtls addi bdnz r0, r3 0, r0, r3 r3, r3, 0x20 LockingLoop void main (void) { int i = 0; } /* /* /* /* Establish Lock that Increment Decrement address line in address CTR, and in cache for 32B cache line of 0's */ cache */ pointer by 32 B */ loop back if it is not yet zero */ /* Dummy idle counter */ cache_enable(); MMU_init_TLB6(); cache_lock_4KB(); /* Invalidate then enable cache in copyback mode */ /* Define 4KB R/W space starting at 0x4004 0000, no transl’n*/ /* Lock 4KB R/W space starting at 0x4004 0000 in cache */ while (1) { i++; } /* Loop forever */ Qorivva Simple Cookbook, Rev. 4 90 Freescale Semiconductor 10 PLL: Initializing System Clock (MPC551x, MPC55xx) 10.1 Description Task: Initialize the system clock to run at 64 MHz from the PLL (FMPLL on MPC555x) whose input is an 8 MHz crystal. Enable, if necessary, CLKOUT. After initializing the PLL predivider, divider, and multiplier, the PLL LOCK is tested by simply polling for the desired status to occur. In a real system, a maximum timeout mechanism would be implemented. If LOCK was not achieved in a maximum time, then you can log an error message and reset the part. CLKOUT has a maximum specified value, hence sysclk has an external bus division factor (EBDF) before sysclk reaches CLKOUT. For this example, 16 MHz CLKOUT will be used. Attempting to increase the frequency in one step may cause overshoot of the PLL beyond the maximum sysclk specification and briefly cause a sharp increase in current demand from the power supply. Therefore two frequency increases are used. The second increase only changes the divider after the PLL feedback loop, which does not cause change of lock because the divider is after the feedback path. Exercise: Measure CLKOUT frequency. Then modify code to produce a new frequency. MPC5500 Multiplier: EMFD (MPC551x) MFD (MPC555x) PLL 8 MHz Crystal PFD / Charge Pumps, Filter, Current Controlled Oscillator Predivider: EPREDIV (MPC551x) PREDIV (MPC555x) Fref Predivider Output Frequency Fsys (sysclk) = 64 MHz Divider: ERFD (MPC551x) RFD (MPC555x) SIU fvco (MPC551x) fico (MPC555x) CLKOUT = 16 MHz EBDF Figure 21. PLL Example Block Diagram (sysclk reset default values: 16 MHz for MPC551x, 12 MHz for MPC555x with 8 MHz crystal, 8 MHz for MPC563x with 8 MHz crystal) Table 37. Signals for PLL Example MPC551x Family Signal CLKOUT eMIOS Ch 12 1 Pin Name SIU PCR No. PE6 70 MPC555x Family Package Pin No. 144 QFP 176 QFP 208 BGA 67 83 P13 Function Name SIU PCR No. CLKOUT – Package Pin No. EVB 496 BGA 416 BGA 324 BGA 208 BGA 144 QFP AF25 AE24 AA20 T141 – (used for test only if no CLKOUT) xPC 563M PJ8–8 Available only on MPC563x 208 BGA. Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 91 10.2 Design Loss of lock and loss of clock detection is not enabled in this example. Because changing the predivider or multiplier can cause loss of lock, the loss-of-lock circuitry and interrupts would not normally be enabled until after these steps are executed. The preliminary specification for MPC5516 has a maximum CLKOUT frequency of 25 MHz, so EBDF will be used to divide the 64 MHz sysclk by 4, producing CLKOUT at 16 MHz. MPC563x note: some packages do not have CLKOUT, so eMIOS channel 12 is configured as OPWFMB to verify sysclk frequency. eMIOS channel 12’s output frequency is sysclk divided by 16. In devices that do not exit reset with PLL enabled as system clock, it is often good practice to initialize PLL registers with the multiplier and dividers before turning on the oscillator. Otherwise the reset default values may try to force the PLL to run beyond its frequency specifications. Be careful not to select a system clock that is not enabled. For example, SIU_SYSCLK[SYSCLKSEL] on MPC551x or FMPLL_ESYNCR1[CLKCFG] on MPC563xM is used to specify the system clock. Do not initialize this field to a source, such as PLL or an internal reference clock, that is not turned on and ready. 10.2.1 MPC551x MPC551x devices have a system clock frequency formula from the PLL of: (EMFD + 16) Fsys = Fref -----------------------------((EPREDIV + 1) (ERFD+1) For an 8 MHz crystal (Fref) and a target frequency of 64 MHz (Fsys), the above formula is used below with values chosen for final MPC551x multiplier and dividers as shown. (See the MPC551x Reference Manual for allowed values for the multiplier and dividers.) Note: Prediv. Output Freq range is 4 MHz to 10 MHz. Fsys (EMFD + 16) (32 + 16) 64 MHz 8 ------ = ------------ = ---- = ------------------ = ----------------------------(0+1) (5+1) (EPREDIV + 1) (ERFD+1) 1 Fref 8 MHz For this MPC551x example we have final values of EMFD = 32, EPREDIV = 0, and ERFD = 5. We must be careful the maximum system clock frequency and VCO minimum/maximum frequencies are within the specification in the Data Sheet as the multiplier and dividers are changed. The steps below show the effect on these frequencies, and use the directions given in the PLL chapter of the MPC5510 Microcontroller Family Reference Manual (June 2007), to accommodate the frequency overshoot. These documented steps are used even though we are using conservative frequencies in this example. Note: ERFD should values, per the reference manual, are mostly odd numbers. Qorivva Simple Cookbook, Rev. 4 92 Freescale Semiconductor Table 38. MPC551x Steps for Programming Enhanced PLL PLL_ESYNCR1 Crystal Freq. Step PLL_ESYNCR Predivider VCO Freq. 2 Output Fsys (192 MHz–680 MHz (3 MHz–66 MHz per Frequency per prelim. Data prelim. Data Sheet) (4 MHz ERFD Sheet) 10 MHz) EPREDI V EMF D 1 83 5 4 MHz (396 MHz) (66 MHz) Write > final ERFD value to PLL_ESYNCR2 (0x000 0007) 1 83 7 4 MHz (396 MHz) (49.5 MHz) Write final EPREDIV and EMFD values to PLL_ESYNCR1 (0xF000 0020) 0 32 7 8 MHz (384 MHz) (45.5 MHz) Turn on OSC and wait for PLL to LOCK – – – – – – Write final ERFD value to PLL_ESYNCR2 (0x0000 0005) 0 32 5 8 MHz 384 MHz 64 MHz 1 83 5 6 MHz (594 MHz) (99 MHz) Write > final ERFD value to PLL_ESYNCR2 (0x000 0005) 1 83 5 6 MHz (594 MHz) (99 MHz) Write final EPREDIV and EMFD values to PLL_ESYNCR1 (0xF002 0030) 2 48 5 4 MHz (256 MHz) (42.6 MHz) Turn on OSC and wait for PLL to LOCK – – – – – – Write final ERFD value to PLL_ESYNCR2 (0x0000 0003) 2 48 3 4 MHz 256 MHz 64 MHz 8 MHz Reset Default Values 12 MHz Reset Default Values 10.2.2 MPC555x MPC555x devices have a system clock frequency formula of: (MFD + 4) Fsys = Fref -----------------------------((PREDIV + 1) 2RFD) For an 8 MHz crystal (Fref) and a target frequency of 64 MHz (Fsys), the above formula is used below with values chosen for MPC555x multiplier and dividers as shown. (See the MPC555x Reference Manual for allowed values for the multiplier and dividers.) Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 93 Fsys (MFD + 4) (12 + 4) 64 MHz 8 ------ = ------------ = ---- = ------------------ = ----------------------------((1 + 1) 20) ((PREDIV + 1) 2RFD) 1 Fref 8 MHz For this example with an 8 MHz crystal, we have final values of MFD = 12, PREDIV = 1, and RFD = 0. If a 40 MHz crystal (Fref) is used, such as on Freescale’s MPC5561 and MPC5567 EVBs, the formula for 64 MHz sysclk is shown below. (See the reference manual for that device to learn the allowed values for multiplier and dividers.) Fsys (MFD + 4) (12 + 4) 64 MHz 8 ------ = ------------ = ---- = ------------------ = ----------------------------((4 + 1) 21) ((PREDIV + 1) 2RFD) 5 Fref 40 MHz For this example with a 40 MHz crystal, we have final values of MFD = 12, PREDIV = 4, and RFD = 1. One must ensure the PLL specifications are not violated for the individual processor’s data sheet. For example, the MPC5554 Microcontroller Data Sheet, Rev 1.4, contains the specifications given here. (Be sure to check your microcontroller’s latest data sheet for actual specified values.) • ICO frequency (fico): 48 MHz to the maximum Fsys. (64 MHz is used here.) • Predivider output frequency: 4 MHz to the maximum frequency. We will use the same pattern of setting sysclk in two stages: the RFD value is initially set to the final RFD value + 1, then wait for LOCK before setting the final RFD value. By waiting to achieve LOCK before increasing to the final frequency, there is the benefit that the transient demand to the circuit board’ s power is reduced and we can eliminate the potential risk of overclocking the CPU and peripherals, which could lead to unpredictable results. CLKOUT is assigned to the pad by default after reset, so no writing to its SIU_PCR is required. However, its SIU_PCR does have controls for disabling the pad’s output buffer and drive strength. 10.2.2.1 MPC563xM Differences MPC563xM implements both a legacy mode, where PLL output frequency uses the MPC555x formula, and an enhanced mode, which uses a different formula offering more frequency choices. This example uses the legacy mode, which is more compatible with existing MPC555x code. Another difference is that the PLL exits reset in the bypass mode, so sysclk operates at the crystal frequency. To have sysclk be based on the PLL output, bit field FMPLL_ESYNCR1[CLKCFG] must be changed. Qorivva Simple Cookbook, Rev. 4 94 Freescale Semiconductor Table 39. MPC551x, MPC555x Steps for PLL Example Pseudo Code Step Relevant Bit Fields Init Assign pad CLKOUT signal (MPC551x only). sysclk • Pad assignment — CLKOUT • Output buffer is enabled • Slew rate control = max. (fastest slew rate) Initialize External Bus Divider Factor from default divide by 2 value to desired divide by 4 value. Desired final CLKOUT = 16 MHz = 64 MHz / (3+1). PA = 1 OBE = 1 SRC = 3 EBDF = 3 MPC551x MPC555x SIU_PCR[70] = 0x060C – SIU_ECCR [EBDF] = 3 Change clock to normal mode with crystal reference. CLKCFG = 7 (MPC551x and MPC563x only) Set initial multiplier and dividers for 64 MHz sysclk: MPC551x with an 8 MHz crystal input, use: • Predivider = 0 + 1 = 1 • Multiple = 32 + 16 = 48 • Divider+1 = 7 + 1 = 8 MPC551x with a 12 MHz crystal input, use: • Predivider = 2 + 1 = 3 • Multiple = 48+ 16 = 4 • Divider+1 = 5 + 1 = 6 MPC555x with an 8 MHz crystal input, use: • Predivider = 0 + 1 = 1 • Multiplier = 12 + 4 = 16 • Divider = 21 = 2 MPC555x with a 40 MHz crystal input, use: • Predivider = 4 +1 = 5 • Multiplier = 12 + 4 = 16 • Divider = 22 = 4 MPC551x: 8 MHz • EPREDIV = 0 • EMFD = 32 (0x20) • ERFD = 7 MPC551x, 12 MHz: • EPREDIV = 2 • EMFD = 48 (0x30) • ERFD = 5 MPC555x, 8 MHz: • PREDIV = 1 • MFD = 12 (0xC) • RFD = 1 MPC555x 40 MHz: • PREDIV = 4 • MFD = 12 (0xC) • RFD = 2 Enable external oscillator (MPC551x only) XOSCEN = 1 [NOTE: Make sure there is a delay such as waiting to lock before changing clock to run on PLL.] MPC563x: FMPLL_ESYNCR1 [CLKCFG] = 7 PLL_ESYNCR2 = 0x0000 0007 for 8 MHz crystal, = 0x0000 0005 for 12 MHz crystal FMPLL_SYNCR = 0x1608 000 for 8 MHz crystal, = 0x4610 0000 for 40 MHz crystal PLL_ESYNCR1 = 0xF000 0020 for 8 MHz crystal, = 0xF002 0030 for 12 MHz crystal CRP_CLKSRC [XOSCEN] = 1 – Wait for PLL_SYNSR [LOCK] = 1 Wait for FMPLL_SYNSR [LOCK] = 1 Wait for PLL to lock Wait for LOCK = 1 Set final divider after feedback loop: MPC551x with an 8 MHz crystal: • Divider = 1 + 5 = 6 MPC551x with a 12 MHz crystal: • Divider = 1 + 3 = 4 MPC555x with 8 MHz crystal: • Divider = 20 = 1 MPC555x with 40 MHz crystal: • Divider = 21 = 2 MPC551x, 8 MHz: • EFRD = 5 MPC551x 12 MHz: • EFRD = 3 MPC555x, 8 MHz: • RFD = 0 MPC555x, 40 MHz: • RFD = 1 PLL_ESYNCR2 FMPLL_SYNCR = 0x0000 0005 = 0x1600 000 for 8 MHz crystal, for 8 MHz crystal, = 0x0000 0003 = 0x4608 0000 for 12 MHz crystal for 40 MHz crystal SYSCLKSEL = 2 SIU_SYSCLK [SYSCLKSEL] = 2 Select PLL for sysclk [and optionally could divide sysclk to peripherals for lower power] (MPC551x only) – Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 95 10.3 10.3.1 /* /* /* /* /* /* /* /* /* /* Code MPC551x with 8 MHz crystal main.c - PLL-sysclk example */ Description: Change sysclk to run from PLL 64 MHz based on 8 MHz crystal. */ Copyright Freescale Semiconductor, Inc 2007. All Rights Reserved */ Rev 1.0 Jul 11 2007 SM- Initial version */ Rev 1.1 Aug 08 2007 SM - Changed sysclk to run at 64 MHz */ Rev 1.2 May 18 2009 SM - Updated to also show 12 MHz crystal implementation */ Rev 1.3 Aug 12 2009 SM - Changed initial ERFD value to be an odd number */ Notes: */ 1. MMU not initialized; must be done by debug scripts or BAM */ 2. SRAM not initialized; must be done by debug scripts or in a crt0 type file */ #include "mpc5510.h" void main (void) { volatile uint32_t i=0; /* Dummy idle counter */ SIU.PCR[70].R = 0x060C; /* Assign pad PortE[6] as CLKOUT signal */ SIU.ECCR.B.EBDF = 3; /* Divide sysclk by 3+1 for CLKOUT */ /* Use 2 of the next 4 lines: */ FMPLL.ESYNCR2.R = 0x00000007; /* 8MHz xtal: ERFD to initial value of 7 */ FMPLL.ESYNCR1.R = 0xF0000020; /* 8MHz xtal: CLKCFG=PLL, EPREDIV=0, EMFD=0x20*/ /*FMPLL.ESYNCR2.R = 0x00000005; */ /* 12MHz xtal: ERFD to initial value of 5 */ /*FMPLL.ESYNCR1.R = 0xF0020030; */ /* 12MHz xtal: CLKCFG=PLL, EPREDIV=2, EMFD=0x30*/ CRP.CLKSRC.B.XOSCEN = 1; /* Enable external oscillator */ while (FMPLL.SYNSR.B.LOCK != 1) {}; /* Wait for PLL to LOCK */ /* Use 1 of the next 2 lines: */ FMPLL.ESYNCR2.R = 0x00000005; /* 8MHz xtal: ERFD change for 64 MHz sysclk */ /*FMPLL.ESYNCR2.R = 0x00000003; */ /* 12MHz xtal: ERFD change for 64 MHz sysclk */ SIU.SYSCLK.B.SYSCLKSEL = 2; /* Select PLL for sysclk */ } while (1) { i++; } 10.3.2 /* /* /* /* /* /* /* /* /* /* Loop forever */ MPC555x with 8 MHz crystal main.c - PLL-sysclk example for MPC555x*/ Description: Set PLL to run at 64 MHz based on 8 MHz crystal */ Copyright Freescale Semiconductor, 2007. All rights reserved. */ Rev 1.0 Jul 12 2007 SM - Initial version */ Rev 1.1 Aug 14 2007 SM - Changed sysclk to 64 MHz */ Rev 1.2 May 5 2008 SM - Added alternate code if 40 MHz crystal is used */ Notes: */ 1. MMU not initialized; must be done by debug scripts or BAM */ 2. L2SRAM not initialized; must be done by debug scripts or in a crt0 type file */ #include "mpc5554.h" void main (void) { volatile uint32_t i=0; /* Dummy idle counter */ SIU.ECCR.B.EBDF = 3; /* Divide sysclk by 3+1 for CLKOUT */ /* For 8 MHz crystal, use the next 3 lines */ FMPLL.SYNCR.R = 0x16080000; /* Initial values: PREDIV=1, MFD=12, RFD=1 */ while (FMPLL.SYNSR.B.LOCK != 1) {}; /* Wait for FMPLL to LOCK */ FMPLL.SYNCR.R = 0x16000000; /* Final value for 64 MHz: RFD=0 */ /* For 40 MHz crystal, use the next 3 lines */ /* FMPLL.SYNCR.R = 0x46100000; *//* Initial values: PREDIV=4, MFD=12, RFD=1 */ /* while (FMPLL.SYNSR.B.LOCK != 1) {}; *//* Wait for FMPLL to LOCK */ /* FMPLL.SYNCR.R = 0x46080000; *//* Final value for 64 MHz: RFD=0 */ } while (1) { i++; } /* Loop forever */ Qorivva Simple Cookbook, Rev. 4 96 Freescale Semiconductor 10.3.3 /* /* /* /* /* /* /* /* /* /* MPC563x with 8 MHz crystal main.c - PLL-sysclk example */ Description: Set PLL to run at 64 MHz based on 8 MHz crystal for MPC563x */ For testing devices without CLKOUT, an eMIOS channel is used */ Copyright Freescale Semiconductor, 2008. All rights reserved. */ Rev 1.0 Jul 12 2007 SM- Initial version */ Rev 1.1 Aug 14 2007 SM -Changed sysclk to 64 MHz */ Rev 1.2 Apr 30 2008 SM- Modified for MPC563x including adding EMIOS OPWFM output*/ Notes: */ 1. MMU not initialized; must be done by debug scripts or BAM */ 2. L2SRAM not initialized; must be done by debug scripts or in a crt0 type file*/ #include "mpc563m.h" /* Used for MPC563m devices */ void initEMIOS(void) { EMIOS.MCR.B.GPRE= EMIOS.MCR.B.ETB = EMIOS.MCR.B.GPREN EMIOS.MCR.B.FRZ = 0x3;/* eMIOS clk= sysclk/(GPRE+1)= sysclk/4 */ 0; /* Ext. time base is disabled; Ch 23 drives ctr bus A */ = 1;/* Enable eMIOS clock */ 0;/* Disable freezing channel counters in debug mode */ } void initEMIOSch12(void) { /* EMIOS CH 12:Output Pulse Width & Freq Modl'n Buf*/ /* Period = 4 emios clks, Duty = 2 eMIOS clks */ /* If 8MHz sysclk, Freq= 8MHz/4/4 = 500Kz (2us per.)*/ /* If 64MHz sysclk, Freq= 64MHz/4/4= 4MHz (250ns per.)*/ EMIOS.CH[12].CBDR.R = 4; /* Period= 4 emios clocks= 16 sysclks */ /* (32usec,4usec for 8M,64Msysclk) */ EMIOS.CH[12].CADR.R = 3; /* Duty cycle in emios clks */ EMIOS.CH[12].CCR.B.UCPRE = 0; /* Channel counter uses divide by (0+1) prescaler */ EMIOS.CH[12].CCR.B.UCPREN = 1; /* Channel counter's prescaler is loaded & enabled*/ EMIOS.CH[12].CCR.B.EDPOL = 1; /* Polarity is active high */ EMIOS.CH[12].CCR.B.MODE= 0x58;/* Mode= 0PWFMB, flag on B match*/ SIU.PCR[191].B.PA = 1; /* Initialize pad for eMIOS channel. */ SIU.PCR[191].B.OBE = 1; /* Initialize pad for output */ } void main (void) { volatile uint32_t i=0; /* Dummy idle counter */ initEMIOS(); /* Init. eMIOS to provide sysclk/4 to eMIOS channels */ initEMIOSch12(); /* Init. eMIOS channel 12 for sysclk/16 OPWFMB */ EMIOS.MCR.B.GTBE = 1; /* Start timers/counters by enabling global time base */ SIU.ECCR.B.EBDF = 3; /* Divide sysclk by 3+1 for CLKOUT */ FMPLL.ESYNCR1.B.CLKCFG = 0X7; /* Change clk to PLL normal mode from crystal */ FMPLL.SYNCR.R = 0x16080000; /* Initial values: PREDIV=1, MFD=12, RFD=1 */ while (FMPLL.SYNSR.B.LOCK != 1) {}; /* Wait for FMPLL to LOCK */ FMPLL.SYNCR.R = 0x16000000; /* Final value for 64 MHz: RFD=0 */ } while (1) { i++; } /* Loop forever */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 97 11 PLL: Initializing System Clock (MPC56xxB/P/S) 11.1 Description Task: Initialize the system clock to run at 64 MHz from the PLL whose input is an 8 MHz or 40 MHz crystal. Enable CLKOUT to observe frequency of 16 MHz IRC, external oscillator (crystal), then FMPLL0. MPC56xxB/P/S mode transition logic simplifies software by not requiring checking the status bits for external oscillator stable and PLL locked. A mode transition is until not complete until: 1. the external oscillator is stable (if the XOSCON bit is set in the targeted mode’s configuration) & 2. the PLL is locked (if the PLLON bit was also set in the targeted mode’s configuration). Exercise: Measure CLKOUT frequency while stepping through code and verify proper frequencies. MPC56xxB/P/S Clock Generation Module 16 MHz System Clock Selector (may have more inputs) IRC Crystal OSC0 FMPLL0 Fsys (sysclk) = 64 MHz Output Clock Selector (may have more inputs) Divide by 1, 2, 4 or 8 Output Clock SIU CLKOUT = 16 MHz Pad Assignment Figure 22. PLL Example Block Diagram Table 40. Signals for PLL Example MPC56xxB Family Port Signal CLKOUT eMIOS Ch 21 PA[0] SIU PCR No. 0 MPC56xxP Family Package Pin No. Port 100 144 176 LQFP LQFP BGA 12 16 G4 PB[6] MPC56xxS Family SIU Package Pin No. PCR 100 144 No. LQFP LQFP 22 96 Port SIU PCR No. Package Pin No. 144 176 LQFP LQFP 208 BGA 138 PH[4] 103 47 61 R5 (used on cut 1 MPC56xxS because cut 1did not have CLKOUT) PA[1] 1 136 166 B1 Qorivva Simple Cookbook, Rev. 4 98 Freescale Semiconductor 11.2 11.2.1 Design Mode Use Mode Transition is required for changing mode entry registers. Hence even enabling the crystal oscillator to be active in the current mode, for example default mode (DRUN,) requires enabling the crystal oscillator in DRUN mode configuration register (ME_DRUN_MC) then initiating a mode transition to the same DRUN mode. This example changes from DRUN mode to RUN0 mode, which is the normal, expected use after power up. This minimal example simply polls a status bit to wait for the targeted mode transition to complete. However, the status bit could instead be enabled to generate an interrupt request (assuming the INTC is initialized beforehand). This would allow software to complete other initialization tasks instead of brute force polling of the status bit. It is normal to use a timer when waiting for a status bit to change. This example by default would have a watchdog timer expire if for some reason the mode transition never completed. One could also loop code on incrementing a software counter to some maximum value as a timeout. If a timeout was reached, then an error condition could be recorded in EEPROM or elsewhere. ce t Table 41. Mode Configurations Summary for MPC56xxB/P/S PLL Example Modes are enabled in ME_ME Register. Settings Mode Mode Config. Register Mode Config. Register Value DRUN ME_DRUN_MC 0x001F 0010 (default) RUN0 Memory Power Mode Clock Sources ME_RUN0_MC 0x001F 0074 sysclk Selection 16MHz IRC XOSC0 PLL0 PLL1 (MPC Data 56xxP/S Flash only) Code Flash Main Voltage Reg. I/O Power Down Ctrl 16 MHz IRC On Off Off Off Normal Normal On Off PLL0 On On On Off Normal Normal On Off Other modes are not used in example It is good practice after a mode transition to verify the desired mode was entered by checking the ME_GS[S_CURRENTMODE] field. If there was a hardware failure, a SAFE mode transition could preempt the desired mode transition. There is an interrupt that, if enabled, can be used generate an interrupt request upon a SAFE mode transition. Peripherals also have configurations to gate clocks on and off, enabling low power. The following table summarizes the peripheral configurations used in this example for MPC56xxB and MPC56xxS. MPC56xxP does not have peripheral configurations for XOSC0 nor PLL0, hence no code is needed for these. Because initial silicon for MPC56xxS did not include CLKOUT, an eMIOS channel is used to indicate final sysclk frequency. MPC56xxP does not require any peripheral clock gating configuration for this example. Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 99 Table 42. Peripheral Configurations for MPC56xxB/P/S PLL Example Low power modes are not used in example. Peripherals Selecting Configuration PeriPeri. Enabled Modes pheral Config. Config. Register RUN3 RUN2 RUN1 RUN0 DRUN SAFE TEST RESET PC1 ME_ RUNPC_ 1 0 0 0 1 0 0 0 0 Peripheral PCTL Reg. # SIUL (MPC56xxB/S) used for enabling pad for output clock 68 Other peripheral configurations are not used in example 11.2.2 PLL Calculations The formula for PLL output clock (also called PHI) in normal mode is: ldf FPLL output clock= Fxtal ----------------(idf odf) which in terms of the PLL’s Control Register field names is equivalent to: FPLL NDIV = F -----------------------------output clock xtal ((IDIF + 1) 2ODIF+1) Also Fvco must be within a range of 256 MHz and 512 MHz. Fvco is defined as: NDIV Fvco = Fxtal --------------(IDIF + 1) The following table summaries calculations to obtain desired 64 MHz frequency for different crystal frequencies. Table 43. MPC56xxB/P/S PLL0 Calculations (NDIV, IDF and ODF are bit field values in GCM FMPLL0 Control Register. Verify range values in microcontroller reference manual. Ranges shown are per MPC5604B Reference Manual, MPC5604B RM, Rev. 1, 4/2008. 11.2.3 Fxtal NDIV Range: 32 - 96 IDF Range: 0 - 14 Fvco Range: 256 - 512 MHz ODF Range: 0-3 FPLL output clock CR value 8 MHz 64 0 512 MHz 2 64 MHz 0x0240 0100 40 MHz 64 4 512 MHz 2 64 MHz 0x1240 0100 Progressive Clock Switching Progressive clock switching is expected to be used on system clock. When changing the system clock to run on a PLL frequency, the PLL locks at a divided frequency, then gradually decreases the division until Qorivva Simple Cookbook, Rev. 4 100 Freescale Semiconductor it is divided by 1. (See illustration below.) The effect is to gradually increase current consumption instead of a single large increase. PLL lock frequency Division factor of 8, 4, 2 then 1 PLL output clock In this example, the system clock changes from 16 MHz internal reference clock to a PLL output clock running at 64 MHz. When the PLL locks, the mode transition can complete. However with progressive clock switching, the PLL output frequency, sysclk in this example, is initially divided by 8 then changes gradually to the programmed frequency without losing lock. The table below illustrates how long this gradual increase takes. Table 44. Progressive Clock Switching for 64 MHz PLL operation PLL lock frequency Division PLL output clock Factor frequency Number of PLL output clock cycles during division Number of Accumulated time PLL lock frequency cycles (for 64 MHz PLL during division lock frequency) 64 MHz 8 8 MHz 8 8 x 8 = 64 1 usec 64 MHz 4 16 MHz 16 4 x 16 = 64 2 usec 64 MHz 2 32 MHz 32 2 x 32 = 64 3 usec 64 MHz 1 64 MHz onward - - As shown above, after the mode transition several microseconds are needed before the PLL output clock, (sysclk in this example) is running at full frequency. The number of system clock cycles before running at full frequency is 8 + 16 + 32 = 56 sysclks, which takes 3 usec. Progressive clock switching is enabled in the Clock Generation Module’s (MC_CGM’s) FMPLL Control Register (CR), en_pll_sw bit. 11.2.4 Clock Monitor Unit (CMU) — XOSC Frequency Monitor One feature of the CMU is monitoring the XOSC frequency with respect to the 16 MHz FIRC. If the crystal frequency is less than FIRC 2RCDIV then, by reset default settings, a reset occurs. RCDIV is a programmable field in the CMU_CSR register. For MPC56xxB/S, RCDIV defaults to a value of three, so as long as FXOSC > 16 MHz 23 (2 MHz) no action takes place. However, MPC56xxP, which may have a 40 MHz crystal for FlexRAY operation, has a reset default of RCDIV = 0. Therefore the threshold is 16 MHz 20 (16 MHz). If, for example, an 8 MHz crystal is used instead of a 40 MHz crystal, CMU_CSR[RCDIV] must change before turning on XOSC in the mode configuration register. Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 101 11.2.5 Oscillator Stabilization Counter Both FIRC and SIRC have counting periods that are used after reset and when the oscillator is switched on. This counting period ensures that the external oscillator clock signal is stable before it can be selected by the system.1 Depending on the crystal, the reset default value in CGM_FXOSC_CTL[EOCV] for FIRC or CGM_SXOSC_CTL[EOCV] for SIRC may need to be changed before the oscillator is turned on. On MPC56xxS “cut 1,” CGM_FXOSC_CTL contained an oscillator enable bit which was disabled after reset. This control has been removed from MPC56xxS, and is also not on the other devices covered here. Table 45. MPC5606B, MPC56xxP, MPC56xxS Steps for PLL Example Relevant Bit Fields Step Init Modes and Clock Enable desired modes Pseudo Code MPC56xxB RUN0, DRUN=1 If necessary, adjust XOSC clock monitor frequency: • 8MHz Crystal: monitor Fxosc > 4 MHz • 40MHz Crystal: monitor Fxosc > 16 MHz MPC56xxP MPC56xxS ME_ME = 0x0000 001D - RCDIV=4 RCDIV=0 40 MHz Crystal: CGM_ CMU_CSR = 0x0000 0000 (default value for MPC56xxP) - 8 MHz Crystal: 40 MHz Crystal: 8 MHz Crystal: CGM_ CGM_ CGM_ FMPLL_CR FMPLL[0]_CR FMPLL[0]_CR = 0x0240 0100 = 0x1240 0100 = 0x0240 0100 Initialize PLL0 dividers to provide 64 MHz for input crystal before mode configuration, using progressive clock switching: • 8MHz Crystal: FMPLL_CR=0x02400100 • 40MHz Crystal: FMPLL_CR=0x12400100 Configure RUN0 Mode: • I/O Output power-down: no safe gating • Main Voltage regulator is on (default) • Data, code flash in normal mode (default) • PLL0 is switched on • Crystal oscillator is switched on • 16 MHz IRC is switched on (default) • Select PLL0 (system pll) as sysclk PDO=0 MVRON=1 DFLAON, CFLAON= 3 PLL0ON=1 XOSC0ON=1 16MHz_IRCON=1 SYSCLK=0x4 ME_RUN0_MC = 0x001F 0070 MPC56xxB/S: • Peri. Config.1: run in RUN0 mode only. RUN0=1 ME_RUN_PC1 = 0x0000 0010 Assign peripheral configuration to peripherals: • SIUL (MPC5500B/S) ME_PCTL68 = 0x01 Initiate software mode transition to RUN0 mode • Mode and key • Mode and inverted key • Wait for transition complete status flag NOTE: if transition does not complete, check status flags such as ME_GS[XOSC] for cause. • Verify current mode is desired mode NOTE: This step ensures a SAFE mode transition did not occur at this point. TARGET_MODE= RUN0 S_TRANS ME_MCTL =0x4000 5AF0 ME_MCTL =0x4000 A50F wait for ME_GS[S_TRANS] = 0 S_ CURRENTMODE verify ME_GS[S_CURRENTMODE] = RUN0 1.EOCV description, Table 3-9, MPC5604B/S Microcontroller Reference Manual, Rev.4 12 Aug 2009 Qorivva Simple Cookbook, Rev. 4 102 Freescale Semiconductor Table 45. MPC5606B, MPC56xxP, MPC56xxS Steps for PLL Example (continued) Relevant Bit Fields Step Initialize Output Clock /4 (CLKOUT goes from about 4 MHz, then to XTAL/4, then to 16 MHz) Disable Watchdog Enable Output Clock to pin EN=1 Select division of output clock by 4 SELDIV = 2 Select 16 MHz IRC as output clock SELCTL = 1 (MPC56xxB) 0 (MPC56xxP/S) Assign output clock (CLKOUT) to pad: • MPC56xxB: PA[0] pad assignmt = alt func 2 • MPC56xxP: PB[6] pad assignmt = alt func 1 • MPC56xxS: PH[4] pad assignmt =option 3 Pseudo Code MPC56xxB SELCTL = 0 (MPC56xxB) 1 (MPC56xxP/S) Select PLL0 as output clock SELCTL = 2 MPC56xxS CGM_OCEN[EN] = 1 CGM_OCDSSC[SELDIV] = 2 CGM_OCDSSC [SELCTL] = 1 SIU_PCR[0] = 0x0800 Select crystal oscillator as output clock CGM_OCDSSC [SELCTL] = 0 • Write keys to clear soft lock bit • Clear watchdog enable bit MPC56xxP WEN = 0 CGM_OCDSSC [SELCTL] = 0 SIU_PCR[22] = 0x0400 SIU_PCR[103] = 0x0E00 CGM_OCDSSC [SELCTL] = 1 CGM_OCDSSC[SELCTL] = 2 SWT_SR = 0x000 0C520 SWT_SR = 0x0000 D928 SWT_CR = 0x8000 010A Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 103 11.3 11.3.1 /* /* /* /* /* /* /* Code MPC56xxB with 8 MHz crystal main.c - PLL example for MPC56xxB */ Description: Set sysclk to FMPLL0 running at 64 MHz & enable CLKOUT */ Jan 14 2009 S Mihalik - Initial version */ May 22 2009 S Mihalik - simplified code */ Jun 24 2009 S Mihalik - Simplified code */ Mar 10 2010 S Mihalik - Modified initModesAndClock & updated header file */ Copyright Freescale Semiconductor, Inc 2009, 2010. All rights reserved. */ #include "MPC5604B_0M27V_0102.h" /* Use proper include file */ void initModesAndClock(void) { ME.MER.R = 0x0000001D; /* Enable DRUN, RUN0, SAFE, RESET modes */ /* Initialize PLL before turning it on: */ /* Use 1 of the next 2 lines depending on crystal frequency: */ CGM.FMPLL_CR.R = 0x02400100; /* 8 MHz xtal: Set PLL0 to 64 MHz */ /*CGM.FMPLL_CR.R = 0x12400100;*/ /* 40 MHz xtal: Set PLL0 to 64 MHz */ ME.RUN[0].R = 0x001F0074; /* RUN0 cfg: 16MHzIRCON,OSC0ON,PLL0ON,syclk=PLL0 */ ME.RUNPC[1].R = 0x00000010; /* Peri. Cfg. 1 settings: only run in RUN0 mode */ ME.PCTL[68].R = 0x01; /* MPC56xxB/S: select ME.RUNPC[1] */ /* Mode Transition to enter RUN0 mode: */ ME.MCTL.R = 0x40005AF0; /* Enter RUN0 Mode & Key */ ME.MCTL.R = 0x4000A50F; /* Enter RUN0 Mode & Inverted Key */ while (ME.GS.B.S_MTRANS == 1) {} /* Wait for mode transition to complete */ /* Notes: */ /* 1. I_TC IRQ could be used here instead of polling */ /* to allow software to complete other init. */ /* 2. A timer could be used to prevent waiting forever.*/ while(ME.GS.B.S_CURRENTMODE != 4){} /* Verify RUN0 is the current mode */ /* Note: This verification ensures a SAFE mode */ /* tranistion did not occur. SW could instead */ /* enable the safe mode tranision interupt */ } void initOutputClock(void) { CGM.OC_EN.B.EN = 1; /* CGM.OCDS_SC.B.SELDIV = 2; /* CGM.OCDS_SC.B.SELCTL = 1; /* SIU.PCR[0].R = 0x0800; /* /* Output Clock enabled (to go to pin) */ Output Clock’s selected division is 2**2 = 4 */ MPC56xxB: Output clock select 16 MHz int RC osc */ MPC56xxB: assign port PA[0] pad to Alt Func 2 */ CLKOUT = 16 MHz IRC/4 = 4MHz */ CGM.OCDS_SC.B.SELCTL = 0; /* MPC56xxB: Assign output clock to XTAL */ /* CLKOUT = Fxtal/4 = 2 or 10 MHz for 8 or 40 MHx XTAL*/ } CGM.OCDS_SC.B.SELCTL = 2; /* Assign output clock to FMPLL[0] */ /* CLKOUT = 64 MHz/4 = 4MHz */ void disableWatchdog(void) { SWT.SR.R = 0x0000c520; /* Write keys to clear soft lock bit */ SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; /* Clear watchdog enable (WEN) */ } void main (void) { vuint32_t i = 0; } /* Dummy idle counter */ initModesAndClock(); /* Initialize mode entries and system clock */ initOutputClock(); /* Initialize Output Clock to 16 M, XOSC, then PLL */ disableWatchdog(); /* Disable watchdog */ while (1) { i++; } Qorivva Simple Cookbook, Rev. 4 104 Freescale Semiconductor 11.3.2 /* /* /* /* /* /* /* MPC56xxP with 40 MHz crystal main.c - PLL example for MPC56xxP */ Description: Set sysclk to FMPLL0 running at 64 MHz & enable CLKOUT */ Jan 14 2009 S. Mihalik - Initial version */ May 11 2009 S. Mihalik - Simplified code */ Jun 24 2009 S. Mihalik - Simplified code */ Mar 10 2010 S Mihalik - Modified initModesAndClock & updated header file */ Copyright Freescale Semiconductor, Inc 2010 All rights reserved. */ #include "Pictus_Header_v1_09.h" /* Use proper include file */ void initModesAndClock(void) { ME.MER.R = 0x0000001D; /* Enable DRUN, RUN0, SAFE, RESET modes */ /* Initialize PLL before turning it on: */ /* Use 2 of the next 4 lines depending on crystal frequency: */ /*CGM.CMU_0_CSR.R = 0x000000004;*/ /*Monitor FXOSC > FIRC/4 (4MHz); no PLL monitor */ /*CGM.FMPLL[0].CR.R = 0x02400100;*/ /* 8 MHz xtal: Set PLL0 to 64 MHz */ CGM.CMU_0_CSR.R = 0x000000000; /* Monitor FXOSC > FIRC/1 (16MHz); no PLL monitor*/ CGM.FMPLL[0].CR.R = 0x12400100; /* 40 MHz xtal: Set PLL0 to 64 MHz */ ME.RUN[0].R = 0x001F0074; /* RUN0 cfg: 16MHzIRCON,OSC0ON,PLL0ON,syclk=PLL0*/ /* Mode Transition to enter RUN0 mode: */ ME.MCTL.R = 0x40005AF0; /* Enter RUN0 Mode & Key */ ME.MCTL.R = 0x4000A50F; /* Enter RUN0 Mode & Inverted Key */ while (ME.GS.B.S_MTRANS == 1) {} /* Wait for mode transition to complete */ /* Notes: */ /* 1. I_TC IRQ could be used here instead of polling */ /* to allow software to complete other init. */ /* 2. A timer could be used to prevent waiting forever*/ while(ME.GS.B.S_CURRENTMODE != 4) {} /* Verify RUN0 is the current mode */ /* Note: This verification ensures a SAFE mode */ /* tranistion did not occur. SW could instead */ /* enable the safe mode tranision interupt */ } void initOutputClock(void) { CGM.OCEN.B.EN = 1; /* Output Clock enabled (to go to pin) */ CGM.OCDSSC.B.SELDIV = 2; /* Output Clock’s selected division is 2**2 = 4 */ CGM.OCDSSC.B.SELCTL = 0; /* MPC56xxP/S: Output clock select 16 MHz int RC osc */ SIU.PCR[22].R = 0x0400; /* MPC56xxP: assign port PB[6] pad to Alt Func 1 */ /* CLKOUT = 16 MHz IRC/4 = 4MHz */ CGM.OCDSSC.B.SELCTL = 1; CGM.OCDSSC.B.SELCTL = 2; } /* MPC56xxP/S: Assign output clock to XTAL */ /* CLKOUT= Fxtal/4 = 2 or 10 MHz for 8 or 40 MHx XTAL*/ /* Assign output clock to FMPLL[0] */ /* CLKOUT = 64 MHz/4 = 4MHz */ void disableWatchdog(void) { SWT.SR.R = 0x0000c520; /* Write keys to clear soft lock bit */ SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; /* Clear watchdog enable (WEN) */ } void main (void) { vuint32_t i = 0; } /* Dummy idle counter */ initModesAndClock(); /* Initialize mode entries and system clock */ initOutputClock(); /* Initialize Output Clock to 16 M, XOSC, then PLL */ disableWatchdog(); /* Disable watchdog */ while (1) { i++; } Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 105 11.3.3 /* /* /* /* /* /* /* MPC56xxS with 8 MHz crystal main.c - PLL example for MPC56xxS */ Description: Set sysclk to FMPLL0 running at 64 MHz & enable CLKOUT */ Jan 14 2009 S Mihalik - Initial version */ May 22 2009 S Mihalik - Simplified code */ Jun 24 2009 S Mihalik - Simplified and verified for cut 2 silicon */ Mar 10 2010 S Mihalik - Modified initModesAndClock & updated header file */ Copyright Freescale Semiconductor, Inc 2010 All rights reserved. */ #include "56xxS_0204.h" /* Use proper include file */ void initModesAndClock(void) { ME.MER.R = 0x0000001D; /* Enable DRUN, RUN0, SAFE, RESET modes */ /* Initialize PLL before turning it on: */ /* Use 1 of the next 2 lines depending on crystal frequency: */ CGM.FMPLL[0].CR.R = 0x02400100; /* 8 MHz xtal: Set PLL0 to 64 MHz */ /*CGM.FMPLL[0].CR.R = 0x12400100;*/ /* 40 MHz xtal: Set PLL0 to 64 MHz */ ME.RUN[0].R = 0x001F0074; /* RUN0 cfg: 16MHzIRCON,OSC0ON,PLL0ON,syclk=PLL0 */ ME.RUNPC[1].R = 0x00000010; /* Peri. Cfg. 1 settings: only run in RUN0 mode */ ME.PCTL[68].R = 0x01; /* MPC56xxB/S: select ME.RUNPC[1] */ /* Mode Transition to enter RUN0 mode: */ ME.MCTL.R = 0x40005AF0; /* Enter RUN0 Mode & Key */ ME.MCTL.R = 0x4000A50F; /* Enter RUN0 Mode & Inverted Key */ while (ME.GS.B.S_MTRANS == 1) {} /* Wait for mode transition to complete */ /* Notes: */ /* 1. I_TC IRQ could be used here instead of polling */ /* to allow software to complete other init. */ /* 2. A timer could be used to prevent waiting forever*/ while(ME.GS.B.S_CURRENTMODE != 4) {} /* Verify RUN0 is the current mode */ /* Note: This verification ensures a SAFE mode */ /* tranistion did not occur. SW could instead */ /* enable the safe mode tranision interupt */ } void initOutputClock(void) { CGM.OC_EN.B.EN = 1; /* CGM.OCDS_SC.B.SELDIV = 2; /* CGM.OCDS_SC.B.SELCTL = 0; /* SIU.PCR[103].R = 0x0E00; /* /* Output Clock enabled (to go to pin) */ Output Clock’s selected division is 2**2 = 4 */ MPC56xxP/S: Output clock select 16 MHz int RC osc */ MPC56xxS: assign port PH[4] pad to Opt 3 (untested)*/ CLKOUT = 16 MHz IRC/4 = 4MHz */ CGM.OCDS_SC.B.SELCTL = 1; /* MPC56xxP/S: Assign output clock to XTAL */ /* CLKOUT = Fxtal/4 = 2 or 10 MHz for 8 or 40 MHx XTAL */ } CGM.OCDS_SC.B.SELCTL = 2; /* Assign output clock to FMPLL[0] */ /* CLKOUT = 64 MHz/4 = 4MHz */ void disableWatchdog(void) { SWT.SR.R = 0x0000c520; /* Write keys to clear soft lock bit */ SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; /* Clear watchdog enable (WEN) */ } void main (void) { vuint32_t i = 0; /* Dummy idle counter */ } initModesAndClock(); /* Initialize mode entries and system clock */ initOutputClock(); /* Initialize Output Clock to 16 M, XOSC, then PLL */ disableWatchdog(); /* Disable watchdog */ while (1) { i++; } Qorivva Simple Cookbook, Rev. 4 106 Freescale Semiconductor 12 FMPLL: Frequency Modulation 12.1 Description Task: Assuming an 8 MHz crystal, initialize the FMPLL to provide an 80 MHz system clock, ramping up in two steps. Enable frequency modulation (FM), with a depth of 1% and a rate of sysclk/40. This example tests the PLL for LOCK and FM status. The program simply waits for the desired status to occur. In real life a maximum timeout mechanism would be implemented. If LOCK was not achieved in a maximum time, then one might log an error message and reset the part. If FM status is not correct then software could loop back and try again, for a selected number of iterations. This implementation is done using C code in the main function for illustration. Normally the PLL would be initialized in assembly language soon after reset, to speed up the rest of initialization. Increasing the PLL frequency produces a brief current demand from the power supply, which varies with the amount of frequency change. This example uses two frequency increases and therefore has lower immediate current increase than if it were done in one step. The second increase changes only the RFD, which does not cause change of lock because the divider is after the feedback path. Exercise: Measure CLKOUT frequency. CLKOUT default frequency is one half of sysclk frequency (Fsys) as defined by SIU_ECCR[EBDF]. Initially CLKOUT will be 1/2 of the 12 MHz reset default frequency (163 ns). CLKOUT at Fsys = 40 MHz is 50 ns; at Fsys = 80 MHz is 25 ns. MPC5500 MFD FMPLL Crystal 8 MHz Fref PREDIV Fsys (sysclk) PFD / Charge Pumps, Filter, Current Controlled Oscillator RFD SIU CLKOUT fico EBDF FM Control Figure 23. FMPLL Example Table 46. Signals for FMPLL Example MPC555x Family Signal Function Name SIU PCR No. CLKOUT CLKOUT – Package Pin No. 496 BGA 416 BGA 324 BGA 208 BGA AF25 AE24 AA20 – Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 107 12.2 Design This example applies to MPC555x devices, but not MPC563x. The formula for system clock is: (MFD + 4) Fsys = Fref -----------------------------((PREDIV + 1) 2RFD) NOTE When using Frequency Modulation, the system frequency plus the 1% or 2% depth must not exceed the device’s rated maximum system frequency. For an 8 MHz crystal (Fref) and a target frequency of 80 MHz (Fsys), derive target values as shown below using the formula for Fsys: Fsys (MFD + 4) (6 + 4) 80 MHz 10 ------ = ------------ = ---- = ------------------ = ----------------------------((0 + 1) 20) ((PREDIV + 1) 2RFD) 1 Fref 8 MHz From this we can see that in this case, for Fsys the target values are: MFD = 6, PREDIV=0, RFD = 0. Initially software will set the RFD to the next higher value (RFD = 1), then it is lowered in a second step to the target frequency. These values provide an ICO frequency (fico) of 80 MHz, which is within the specification of the MPC5554 Data Sheet, Rev 1.4. Now the bit field values for Frequency Modulation can be determined. The percent depth (P) is 1. For a target MFD of 6, M = 480. The expected difference value (EXP) is calculated per the formula in the MPC5554 Reference Manual, section 11.4.4.1: (MFD + 4) M P (6 + 4) 480 1 EXP = -------------------------- = ----------------------- = 48 100 100 Loss of lock and loss of clock detection is not enabled in this example. The PLL will not exit reset until it has locked. Because changing PREDIV or MFD, or enabling FM, can cause loss of lock, the loss-of-lock circuitry and interrupts would not be enabled until after these steps. Qorivva Simple Cookbook, Rev. 4 108 Freescale Semiconductor Table 47. FMPLL: Frequency Modulation Design Steps Step Relevant Bit Fields 1. Initialize PLL for less than desired frequency (40 MHz). • EXP value based on calculation = 48 (0x30) • Multiplication Factor Divider = 6 • PREDIVider = 0 • Initial Reduced Frequency Divider = 1 • Keep Frequency Modulation disabled for now EXP = 0x30 MFD = 6 PREDIV = 0 (default) RFD = 1 DEPTH = 0 (default) 2. Wait for PLL to lock. wait for LOCK = 1 3. Enable FM: • Set DEPTH to 1% • Set RATE DEPTH = 1 RATE = 1 (Fref / 40) 4. Wait for PLL to re-lock (relocks after CALDONE = 1). wait for LOCK = 1 5. Verify calibration completed and successful. CALDONE = 1 CALPASS = 1 6. Go to target frequency (80 MHz). RFD = 0 12.3 /* /* /* /* /* /* Pseudo Code FMPLL_SYNCR = 0x0308 0030 wait for FMPLL_SYNSR[LOCK] = 1 FMPLL_SYNCR = 0x0308 0430 wait for FMPLL_SYNSR[LOCK] = 1 if ((FMPLL_SYNSR[CALDONE != 1) or (FMPLL_SYNSR[CALPASS] != 1)) then log an error FMPLL_SYNCR = 0x0300 0430 Code main.c - FMPLL example */ Rev 1.0 Sept 21 2004 S. Mihalik Copyright Freescale Semiconductor, Inc. 2004 All rights reserved. */ Notes: */ 1. ECC in L2SRAM is not initialized here; must be done by debug scripts 2. Cache is not used */ */ #include "mpc5554.h" void errorLog (void){ while (1) {} } /* Placeholder for FMPLL error - wait forever */ void initFMPLL(void) { FMPLL.SYNCR.R = 0x03080030; /* Initial setting: 40 MHz for 8 MHz crystal*/ while (FMPLL.SYNSR.B.LOCK != 1) {}; /* Wait for LOCK = 1 */ FMPLL.SYNCR.R = 0x03080430; /* Enable FM */ while (FMPLL.SYNSR.B.LOCK != 1) {}; /* Wait for FMPLL to LOCK again */ if ((FMPLL.SYNSR.B.CALDONE != 1) | (FMPLL.SYNSR.B.CALPASS != 1)) { errorLog(); /* Error if calibration is not done or did not pass */ } FMPLL.SYNCR.R = 0x03000430; /* Final setting: 80 MHz for 8 MHz crystal */ } void main (void){ int i=0; /* Dummy counter */ initFMPLL(); /* Initialize FMPLL for 80 MHz & 1% Frequency Modulation depth */ while (1) {i++; } /* Wait forever */ } Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 109 13 Modes: Low Power (MPC56xxB/S) 13.1 Description Task: Using lowest power modes, pulse an output pin to a 10% “on” duty cycle approximately every second (~0.9 seconds off, then ~0.1 seconds on). Repeat until a pin transition occurs. There are two wakeup event timer choices: Autonomous Periodic Interrupt (API) for multiple regular (periodic) short timeouts, and Real Time Counter (RTC), a single longer timeout. This example uses RTC. In STANDBY mode, outputs other than wakeup inputs with enabled pullups are always in high impedance state. Therefore, STANDBY should be used for the longer of the two wakeup timeouts (WT1 in this example) with possibly either a pullup or pulldown resistor externally connected to acheive the desired ON or OFF output state such as to an LED. After the first wakeup timeout (WT1), STANDBY mode exits and the processor goes to the RESET vector or 0x4000 0000 in SRAM if RGM_STDBY[BOOTFROM_BKP_RAM] is set. At this point, the processor is in DRUN mode and software should verify that the reason for entering DRUN was the desired wakeup event. Software then turns on the LED and sets up the new wakeup timeout (WT2). However, because the new output state must be maintained at the pad, you cannot go back to STANDBY, because the pad would enter high impedance and be pulled up or down depending on the external resistor. Therefore STOP mode is used instead of STANDBY for this second shorter wake up timeout (WT2). Software will configure STOP mode to maintain the I/O pin states (using MC_STOP_MC[PDO] field), then enter STOP mode from a RUNx mode. After the WT2 wakeup timeout, STOP mode exits back to the previous RUNx mode with the instruction pointer unchanged. Software here simply re-initializes the wakeup timer to the first value and then re-enters STANDBY mode. By entering STANDBY, the output goes to the high impedance state so the pad is pulled back up or down according to the circuit. Exercise: Change timeouts for a new period and duty cycle. Figure 24. Modes and Wakeup Timeouts WT1 & WT2 DRUN RUN3 RUN3 RUN3 DRUN Current (approx. relative levels) DRUN STOP STANDBY WT1 DRUN pin transition wakeup event STOP STANDBY WT2 WT1 WT2 WT1 Time Qorivva Simple Cookbook, Rev. 4 110 Freescale Semiconductor 13.2 Design 13.2.1 Modes vs. Powered Domains vs. Modules Hardware is used to control switching power on and off to primary power domains (PD0 and PD1) based on the mode being entered to conserve current. When a domain is powered off, then of course all initialization or other data is lost for registers and memory in that domain. Unlike power domains PD0 and PD1, power for PD2 is configurable by software writing to the PCU_PCONF2 register. From the table below we can see that going to STANDBY mode cuts power to core, most peripherals, mode entry configurations, clocks, etc. Hence we need to re-initialize registers in PD0 and PD1 when exiting STANDBY mode, but not other modes such as STOP mode. Table 48. Summary of Powered Modules by Mode and Domain for MPC56xxB/S Mode Entry Low Power example (See chapters in reference manual on “Power Control Unit and Voltage Regulators” and “Power Supplies”) Power Domain PD0 Power Domain PD1 Power Domain PD2 Configuration Register PCU_PCONF0 Configuration Register PCU_PCONF1 Configuration Register PCU_PCONF2 Mode Domain Modules: Domain Modules: PCU, RGM, WKPU, SIRC, MC_ME1, MC_CGM, SIUL, FIRC, CAN Sampler, Core, Flash, FMPLL, other 8K SRAM, peripherals ME_DRUN_MC register 1 13.2.2 Domain Modules: Additional SRAM DRUN Powered Powered Configurable RUN3 Powered Powered Configurable STOP Powered Powered Configurable STANDBY Powered Not Powered Configurable ME module exception: ME_DRUN_MC register is powered during STANDBY. RTC/API Clock Selection and Configuration One of three clock sources below can be selected for the RTC counter clock source: • • • • 128 kHz SIRC (Slow Internal Reference Clock) with optional divider in CGM_SIRC_CTL 16 MHz FIRC (Fast Internal Reference Clock) with optional divder in CGM_FIRC_CTL 32 kHz SXOSC (“Slow” External Oscillator) MPC56xxS only: 4–16 MHz FXOSC (“Fast” External Oscillator) The selected clock can be also divided by 512 and/or 32 in RTC_RTCC before clocking the 32-bit RTC counter. (See RTC/API block diagram in the device reference manual chapter, “Real Time Clock / Autonomous Periodic Interrupt.”) There are two timeouts that can occur and each can generate a wakeup event or interrupt request. Each timeout compares a programmable value to the 32-bit RTC counter. • RTCVAL: for longer timeouts, is always enabled when RTC counter is running Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 111 • APIVAL: for shorter regular timeouts, must be enabled separately The following table provides useful calculations for various clock sources and dividers for RTC and API timeouts. Once we know our wakeup timeout(s), we can use this table to decide on a clock source, whether to use API or RTC wakeup, and whether to use the 512 and/or 32 clock dividers. Our target wakeup timeouts are WT1= 0.9 seconds and WT2 = 0.1 seconds. We will use the 128 kHz SIRC divided by 4 (32 kHz) for a clock source without other dividers. Wakekup source is RTC, since the range of min to max RTC timeouts meets our criteria of 0.1 and 0.9 seconds and SIRC uses less power than FIRC. Note that the RTC Counter Register (RTCCNT) only clears on Power On Reset. Other resets do not change its value. When using the RTC, set the RTC value in RTC_RTCC[RTCVAL] to its current RTCVAL plus an offset, where the offset is the next desired delay in terms of appropriate RTCCNT units. Table 49. RTC/API Clock, Divider, and Timeout Examples for MPC56xxB/S RTC Counter RTC Counter Min. RTC Max. RTC RTC Rollover Min. API Max. API Input Clock Input Clock Timeout Timeout Timeout Timeout Timeout Clock Frequency Period (2 10 ((2 22 – 1) x (2 32 (1 ((2 10 – 1) x RTC_ RTC_ Source (1 / (RTC RTC Counter RTC Counter RTC Counter RTC Counter RTC Counter RTCC RTCC input clock input clock input clock input clock Counter Input input clock [div512] [div32] period) period) period) period) Clock Freq.)) period) div512 div 32 128 kHz SIRC1 with SIRCDIV =0 16 MHz FIRC2 (125 x SIRC) 32 kHz SXOSC3 or 128 kHz SIRC/44 5 FXOSC 1 2 3 4 5 0 0 128 kHz ~7.8 sec ~7.8 sec ~ 8 msec ~ 8 msec ~31 sec ~8.9 hrs 0 1 4 kHz 250 sec 250 sec ~ 256 msec 256 msec ~17 min ~12 days 1 0 250 Hz 4 msec 4 msec ~ 4.09 sec ~4.10 sec ~4.4 hrs ~6.3 mo 1 1 7.8125 Hz 128 msec 128 msec ~131 sec ~131 sec ~6 days ~17 yrs 0 0 16 MHz 62.5 nsec 62.5 nsec ~64 sec 64 sec 0.25 sec ~4.3 min 0 1 500 kHz 2 sec 2 sec ~2 msec 2.048 msec 8 sec ~2.8 hrs 1 0 31.25 kHz 32 sec 32 sec ~33 msec ~33 msec ~2.1 min ~1.5 days 1 1 ~977 Hz 1.024 msec 1.024 msec ~1048 sec ~1049 sec ~1.1 hrs ~1.6 mo 0 0 32 kHz 31.25 sec 31.25 sec ~32 msec 32 msec ~2.1 min ~1.5 days 0 1 1 kHz 1 msec 1 msec 1.023 sec 1.024 sec ~1.1 hrs ~1.6 mo 1 0 62.5 Hz 16 msec 16 msec ~16 sec ~16 sec ~18 hrs ~2.1 yrs 1 1 ~2 Hz 512 msec 512 msec ~524 sec ~524 sec ~24 days ~67 yrs 0 0 8 MHz 125 nsec 125 nsec ~128 sec ~128 sec 0.5 sec ~8.6 min 0 1 250 kHz 4 sec 4 sec ~4.1 msec ~4.1 msec 16 sec ~4.6 hrs 1 0 15.625 kHz 64 sec 64 sec ~65 msec ~66 msec ~4.3 min ~3 days 1 1 ~488 Hz 2.048 msec 2.048 msec ~2.1 sec ~2.1 sec ~2.3 hr ~3.2 mo SIRC is divided by 4 to 32 kHz using reset default value in CGM_SIRC_CTL[SIRCDIV] for MPC56xxB, or GGM_LPRC_CTL[LPRCDIV] for MPC56xxS. FIRC is divided by 1 using reset default value in CGM_RC_CTL[RCDIV] for MPC56xxB, MPC56xxS. 32 kHz SXOSC not available in STANDBY mode. SIRC is divided by 4 to 32 kHz using reset default value in CGM_SIRC_CTL[SIRCDIV] for MPC56xxB, or GGM_LPRC_CTL[LPRCDIV] for MPC56xxS. FXOSC as RTC/API option avaialble only in MPC56xxS. Also, not available in STANDBY mode. Qorivva Simple Cookbook, Rev. 4 112 Freescale Semiconductor 13.2.3 Pin Transition Selection and Configuration One or more pins can be selected for generating a wakeup per the following table. Table 50. Wakeup Sources (Not all ports are available in all packages — see device reference manual) MPC560xS MPC560xB Wakeup Number Port SIU_PCR # Wakeup IRQ to INTC Port SIU_PCR Wakeup IRQ to INTC WKUP0 API n.a. WakeUp_IRQ_0 PA0 PCR0 WakeUp_IRQ_0 WKUP1 RTC n.a. WakeUp_IRQ_0 PB1 PCR17 WakeUp_IRQ_0 WKUP2 PA1 PCR1 WakeUp_IRQ_0 PB3 PCR19 WakeUp_IRQ_0 WKUP3 PA2 PCR2 WakeUp_IRQ_0 PB4 PCR20 WakeUp_IRQ_0 WKUP4 PB1 PCR17 WakeUp_IRQ_0 PB9 PCR25 WakeUp_IRQ_0 WKUP5 PC11 PCR43 WakeUp_IRQ_0 PB10 PCR26 WakeUp_IRQ_0 WKUP6 PE0 PCR64 WakeUp_IRQ_0 PB12 PCR28 WakeUp_IRQ_0 WKUP7 PE9 PCR73 WakeUp_IRQ_0 PC0 PCR30 WakeUp_IRQ_1 WKUP8 PB10 PCR26 WakeUp_IRQ_1 PC10 PCR40 WakeUp_IRQ_1 WKUP9 PA4 PCR4 WakeUp_IRQ_1 PF0 PCR70 WakeUp_IRQ_1 WKUP10 PA15 PCR15 WakeUp_IRQ_1 PF2 PCR72 WakeUp_IRQ_1 WKUP11 PB3 PCR19 WakeUp_IRQ_1 PF3 PCR73 WakeUp_IRQ_1 WKUP12 PC7 PCR39 WakeUp_IRQ_1 PF5 PCR75 WakeUp_IRQ_1 WKUP13 PC9 PCR41 WakeUp_IRQ_1 PF6 PCR76 WakeUp_IRQ_1 WKUP14 PE11 PCR75 WakeUp_IRQ_1 PF8 PCR78 WakeUp_IRQ_2 WKUP15 PF11 PCR91 WakeUp_IRQ_1 PF11 PCR81 WakeUp_IRQ_2 WKUP16 PF13 PCR93 WakeUp_IRQ_2 PF13 PCR83 WakeUp_IRQ_2 WKUP17 PG3 PCR99 WakeUp_IRQ_2 PJ4 PCR108 WakeUp_IRQ_2 WKUP18 PG5 PCR101 WakeUp_IRQ_2 PJ6 PCR111 WakeUp_IRQ_2 WKUP19 PA0 PCR0 WakeUp_IRQ_2 API n.a. WakeUp_IRQ_2 WKUP20 (none) (none) (none) RTC n.a. WakeUp_IRQ_2 These ports can be configured for rising and/or falling edge detection. An analog filter can be enabled in WKUP_WIFER which prevents glitches on those pins. When the filter is enabled, input pulses less than 40 ns (WFI) are filtered (rejected), greater than 1000 ns (WNFI) are not filtered, and those with widths in between may or may not get filtered1. 1. Reference: MPC5604BC Data Sheet, Rev. 5, November 2009, Table 12, “I/O Input DC electrical characteristics, WFI and WNFI parameters.” Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 113 13.2.4 STANDBY Wakeup: Execute from Flash or RAM? When exiting STANDBY mode by pin transition or timer wakeup, there are two options for code execution depending on the configuration of RGM_STDBY[BOOT_FROM_BKP_RAM]: • System boots from flash on STANDBY exit (reset default) • System boots from backup RAM on STANDBY exit Configuring system boot from backup RAM has two advantages: 1. Enables lower power consumption. Reason: the flash does not need to be fully powered. If booting from SRAM and writing C code, there needs to be an initialization (such as initializing the stack pointer and small data areas), which is part of a normal compiler initialization on reset. However, if booting from SRAM and only writing a small amount of code, this could be written in assembly and have a shorter boot sequence. 2. Faster startup time. Reason: flash does not need to be power sequenced. The amount of power savings when booting from RAM and STANDBY wakeup varies by application. This example does not use the feature to boot from backup RAM on standby wakeup, but boots from flash after reset and BAM code execution. 13.2.5 Reset and Wakeup Strategy Summary After reset, the processor is in DRUN mode and software must determine the cause of reset. This example is implemented using the RTC and takes the key actions below based on the following reset causes: • Reset event: Configure RTC wakeup of WT1 duration and enter STANDBY mode (outputs are in high impedance in STANDBY mode). • Wakeup event: Set an output pin, configure RTC wakeup of WT2 duration and enter STOP mode. On STOP mode exit (from RTC timeout of WT2), configure RTC wakeup of WT1 and enter STANDBY mode (outputs return to high impedance in STANDBY mode). • Pin wakeup event (final wakeup for this example): wait forever (do not enter more low power modes). Wakeup from STOP mode exits to the RUNx MODE that was used previously, so instruction execution resumes in that mode from the next instruction, without any reset. Alternative implementations could be based on using the API. Two options are: 1. STANDBY exit with API 1/10th second wakeup Simply increment an API wakeup counter in SRAM that is kept alive (in Power domain 0). At the 9th wakeup, go into STOP mode with the desired pad powered. On the next API wakeup, reset the counter and go back to STANDBY. 2. STANDBY exit with API 1 second wakeup On API wakeup from STANDBY, power the desired pad, set up a 1/10th second PIT interrupt and execute a WAIT instruction (instead of going back to a low power mode). At the PIT interrupt go back to STANDBY and wait for STANDBY wakeup. Qorivva Simple Cookbook, Rev. 4 114 Freescale Semiconductor NOTE Some MPC56xx device documentation list more than one STOP0, HALT0, or STANDBY0 mode. The devices for this example only have one such mode each, so the “0” suffix is not used in this write up, but must be used for coding register names if applicable. NOTE Although STANDBY exit causes a reset, not all registers are reset as in a power-up reset. Power domain 0 registers are not reset, including: RGM registers, ME _DRUN_MC register and PCU registers, RTC registers, CAN SAMPLER registers, etc. 13.2.6 Debugging with Low Power Modes By default, low power modes shut off as much power as possible, including to the debugger connection. To avoid the debugger losing connection, and therefore possibly asserting a reset, the Nexus Development Interface (NDI) module contains features to maintain a debugger connection. Debuggers may or may not implement this feature. The controls to maintain the debugger connection are in the Port Configuration Register (PCR) of the Nexus Development Interface module. These registers are not memory mapped, and are only accessible through the debugger. Users should use appropriate debugger scripts to configure these settings. To maintain a debugger connection, the debugger should set the following: • PCR[LP_DBG_EN] Low Power Debug Enable: Enables debug functionality to support exit from any low power mode. Debug functionality is immediately available for breakpoints, etc. When enabled, low power exit takes longer because the state machine adds steps to wait until after a handshake to the debugger and the debugger response to the handshake completes. • PCR[MCKO_EN] MCKO Enable: Enables clock which will be used in debugger connection. • PCR[MCKO_DIV] MCKO Divistion Factor: Determines MCK0 relative to system clock frequency. Options are SYSCLK divided by 1 (default), 2, 4, or 8. 13.2.7 Mode Use This example simply polls a status bit to wait for the targeted mode transition to complete. However, the status bit could instead be enabled to generate an interrupt request (assuming the INTC is intialized beforehand). This would allow software to complete other intialization tasks instead of brute force polling of the status bit. It is normal to use a timer when waiting for a status bit to change. This example by default would have a watchdog timer expire if for some reason the mode transition never completes. One could also loop code on incrementing a software counter to some maximum value as a timeout. If a timeout was reached, then an error condition could be recorded in EEPROM or elsewhere. Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 115 Table 51. Mode Configurations Summary for MPC56xxB/S Mode Entry Low Power Example Modes are enabled in ME_ME register Settings Mode & Main Use DRUN Mode Config. Register Memory Power Mode Clock Sources Mode Config. Register Value sysclk Selection PLL1 (MPC 56xxP/ S only) 16 MHz XOSC0 PLL IRC 0 Data Flash I/O Main Power Voltage Down Code Reg. Ctrl Flash ME_DRUN_MC 0x001F 0010 (default) 16 MHz IRC On Off Off Off Normal Normal On Off RUN3: ME_RUN3_MC 0x001F 0010 Wakeup 16 MHz IRC On Off Off Off Pwr Dn Normal On Off STOP: ME_STOP_MC 0x0005 000F Hold Pin States: Disabled Off Off Off — Pwr Dn Pwr Dn Off Off STAND- ME_STANDBY 0x0085 000F BY0 _MC Disabled Off — — — Pwr Dn Pwr Dn Off On Other modes are not used in example. Peripherals also have configurations to gate clocks on and off, enabling low power. The following table summarizes the peripheral configurations used here. Table 52. Run Peripheral Configurations for MPC56xxB/S Mode Entry Low Power Example PeriEnabled Modes pheral Peri. Config. Config. & Main Register RUN3 RUN2 RUN1 RUN0 DRUN SAFE TEST RESET Use PC7: Low Pwr Wake up ME_ RUNPC_ 7 1 0 0 0 1 0 0 0 Peripherals Selecting Configuration Peripheral PCTL Reg. # SIUL WKPU RTC/API 68 69 91 Other peripheral configurations are not used in example. Table 53. Low Power Peripheral Configurations for MPC56xxB/S Mode Entry Low Power Example PeriPeri. Enabled Modes pheral Config. STOP Config. Register STANDBY0 PC7: Low Pwr I/O ME_ LPPC_ 7 0 1 Peripherals Selecting Configuration HALT 0 Peripheral SIUL WKPU RTC/API PCTL Reg. # 68 69 91 Other peripheral configurations are not used in example. Qorivva Simple Cookbook, Rev. 4 116 Freescale Semiconductor 13.2.8 Steps and Pseudo Code Table 54. Initial Steps for MPC56xxB/S Mode Entry Low Power Example Disable • Write keys to clear soft lock bit Watchdog • Clear watchdog enable bit Init Modes and sysclk Pseudo Code Relevant Bit Fields Step MPC56xxB SWT_SR = 0x000 0C520 SWT_SR = 0x0000 D928 SWT_CR = 0x8000 010A WEN = 0 Enable desired modes: DRUN, RUN3, STANDBY, STOP ME_ME = 0x0000 248D Configure DRUN, RUN3 Modes: • I/O Output power-down: outputs not disabled PDO=0 • Main Voltage regulator is on MVRON=1 • Data, code flash in normal mode DFLAON = 3, CFLAON = 3 • PLL0 is switched off PLL0ON=0 • Crystal oscilator is switched off XOSC0ON=0 • 16 MHz IRC is switched on (default) 16MHz_IRCON=1 • Select FIRC (16 MHz) as sysclk SYSCLK=0x0 ME_DRUN_MC = 0x001F 0010 ME_RUN3_MC = 0x001F 0010 Configure STOP Mode: • I/O Output power-down: outputs not disabled PDO=0 • Main Voltage regulator is off MVRON=o • Data, code flash in power-down mode DFLAON=1, CFLAON=1 • 16 MHz IRC is switched off 16MHz_IRCON=0 • sysclk is disabled SYSCLK=0xF Configure STANDBY Mode: • Main Voltage regulator is (always) off • 16 MHz IRC is switched off • sysclk is (always) disabled MVRON=0 16MHz_IRCON=0 SYSCLK=0xF ME_STOP0_MC = 0x0005 000F ME_STOP_MC = 0x0005 000F ME_STANDBY0_MC = 0x0085 000F ME_STANDBY_MC = 0x0085 000F Peripheral Configurations: • Run Peri. Cfg. 7: run in DRUN, RUN3 modes DRUN, RUN3 = 1 • Low Pwr Peri. Cfg. 7: run in STOP STOP = 1 ME_RUN_PC7 = 0x0000 0088 ME_LP_PC7 = 0x0000 0400 Assign peripheral configuration to peripherals: • SIUL: select ME_RUN_PC7, ME_LP_PC7 RUN_CFG = 7 • WKPU: select ME_RUN_PC7, ME_LP_PC7 LP_CFG = 7 • RTC/API: select ME_RUN_PC7, ME_LP_PC7 If necessary, wait for mode transition from STANDBY to complete. MPC56xxS ME_PCTL68 = 0x3F ME_PCTL69 = 0x3F ME_PCTL91 = 0x3F S_MTRANS Initiate software mode transition to same mode to enable PCTLs, RUN_PC, LP_PC to latch • Mode and key TARGETMODE = • Mode and inverted key RUN3 • Wait for mode transition complete status flag S_MTRANS NOTE: if transition does not complete, check status flags such as ME_GS[XOSC] • Verify desired target mode was entered while MS_GS[S_MTRANS] = 1 ME_MCTL =0x7000 5AF0 ME_MCTL =0x7000 A50F wait for ME_GS[S_MTRANS] = 1 verify ME_GS[S_CURRENT_MODE] = RUN3 Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 117 Table 54. Initial Steps for MPC56xxB/S Mode Entry Low Power Example (continued) Reset Source Determination and Action Pseudo Code Relevant Bit Fields Step MPC56xxB MPC56xxS If destructive reset event, call function: ResetEvent RGM_DES flags if (RGM_DES != 0) ResetEvent If functional reset event, call function: ResetEvent RGM_FES flags if (RGM_FES != 0) ResetEvent If RTC wakeup, 56xxB: EIF1 56xxS: EIF20 call function: RTCWakeup function If pin transition wakeup from: • MPC56xxB: PE[0] • MPC56xxS: PA[0] wait forever 56xxB: EIF6 56xxS: EIF0 if WKUP_WISR[EIF1] RTCWakekupEvent if WKUP_WISR[EIF20] RTCWakekupEvent if WKUP_WISR[EIF6] if WKUP_WISR[EIF0] while 1 while 1 else wait forever (should never get here!) while 1 Qorivva Simple Cookbook, Rev. 4 118 Freescale Semiconductor Table 55. ResetEvent Function Steps for MPC56xxB/S Mode Entry Low Power Example Pseudo Code Step Init Wakeup Events Relevant Bit Fields Initialize SIRC clock used for wakeup: divide 128 kHz clock by 4 SIRCON_STDBY=1 SIRCDIV = 3 Clear RTC and enable writing to. RTCVAL or APIVAL CNTEN = 0 Initialize RTC wakeup timer • Enable RTC counter • Enable counter freeze during debug • Select clock source 128 kHz SIRC / 4 • Initialize initial RTCVAL to WT1 of 0.9 sec RTCVAL = 900 msec 1 count/33 msec = ~27 MPC56xxB MPC56xxS CGM_SIRC_CTL = 0x0000 0301 CGM_LPRC_CTL = 0x0000 0300 RTC_RTCC = 0x0000 0000 RTC_RTCC = 0xA01B 1000 CNTEN = 1 FRZEN = 1 CLKSEL =1 RTCVAL = 27 Enable Wakeup Rising Edge Event from RTC and: • MPC56xxB: Pin PE[0] • MPC56xxS: Pin PA[0] WKUP._WIREER WKUP._WIREER = = 0x0000 0002 0x0010 0001 Enable analog filter for pad: • MPC56xxB: Pin PE[0] • MPC56xxS: Pin PA[0] WKUP_WIFER = 0x0000 0002 WKUP_WIFER = 0x0000 0001 Enable wakeup event rom RTC and: • MPC56xxB: Pin PE[0] • MPC56xxS: Pin PA[0] WKUP_WRER = 0x0000 0002 WKUP_WRER = 0x0010 0001 WKUP_WIPUER =0x000F FFFF WKUP_WIPUER =0x001F FFFF WKUP_WISR = 0x0000 FFFF WKUP_WISR = 0x0001 FFFF Enable WKUP pin pullups to minimize leakage Clear flags Clear all prior wakeup flags, if any Clear reset’s destructive & functional event flags Transision Initiate software mode transition to STANDBY mode to • Mode and key TARGET_MODE= STANDBY • Mode and inverted key STANDBY Wait for low power mode transition to complete Note: Alternative is to implement a timeout interrupt during a mode instruction, then use WAIT instruction clear RGM_DES & RGM_FES flags ME_MCTL = 0xD000 5AF0 ME_MCTL = 0xD000 A50F while 1 Processor now in STANDY mode Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 119 Table 56. RTC Wakeup Event Function Steps for MPC56xxB/S Mode Entry Low Power Example1 Pseudo Code Step Relevant Bit Fields MPC56xxB Pad config. Configure pad GPIO68 to output 0 to turn on LED • GPIO68 on 56xxB EVB: LED 1, port PE[4] • GPIO68 on 56xxS EVB: LED 3, port PE[6] SIU_GPDO[68]= 0 SIU_PCR[68] = 0x0200 Configure pad as input (to be used for wakeup): • MPC56xxB: Port E[0] • MPC56xxS: Port PA[0]. Can be wired to EVB key Re-init RTC wakeup Clear RTC and enable writing to. RTCVAL or APIVAL. Initialize RTC wakeup timer • Enable RTC counter • Enable counter freeze during debug • Select clock source 128 kHz SIRC / 4 • Initialize initial RTCVAL to WT1 of 0.9sec RTCVAL = 900 msec x 1 count/33 msec = ~27 MPC56xxS SIU_PCR[64] = 0x0103 CNTEN = 0 SIU_PCR[0] = 0x0103 RTC_RTCC = 0x0000 0000 RTC_RTCC = 0xA01B 1000 CNTEN = 1 FRZEN = 1 CLKSEL =1 RTCVAL =0x1B Clear flag Clear RTC wakeup event flag (write 1 to clear) WKUP_WISR = 0x0000 0002 Transition to STOP mode Initiate software mode transition to STOP mode • Mode and key TARGET_MODE= • Mode and inverted key STOP • Wait for STOP mode transtion to complete (could have timeout interrupt for timeout.) WKUP_WISR [= 0x0010 0000 ME_MCTL = 0xA000 5AF0 = 0xA000 A50F Wait for ME_GS[S_MTRANS] STOP mode active. Code continues on wakeup from STOP mode back into same RUN3 mode. Verify RUN Verify mode transition from STOP mode completed mode back to RUN3 mode Re-init RTC wakeup Clear flag Verify ME_GS[CURRENTMODE] = RUN3 Clear RTC and enable writing to RTCVAL or APIVAL CNTEN = 0 RTC_RTCC = 0x0000 0000 Initialize RTC wakeup timer • Enable RTC counter • Enable counter freeze during debug • Select clock source 128 kHz SIRC / 4 • Initialize initial RTCVAL to WT1 of 0.9sec RTCVAL = 900 msec 1 count/33 msec = ~10 RTC_RTCC = 0xA01B 1000 CNTEN = 1 FRZEN = 1 CLKSEL =1 RTCVAL = 10 Clear RTC wakeup event flag WKUP_WISR = 0x0000 0002 Transision Initiate software mode transition to STANDBY mode to • Mode and key TARGET_MODE= STANDBY • Mode and inverted key STANDBY mode Wait to enter STANDBY (could have timeout) WKUP_WISR [= 0x0010 0000 ME_MCTL = 0xD000 5AF0 = 0xD000 A50F while 1 STANDBY mode active. On exit, reset vector is taken. 1 After RTC wakeup timeout, software outputs a logic signal to turn on an Evaluation Board (EVB) LED, configures next wakeup timeout, and transitions to STOP mode. Upon wakeup timeout from STOP mode, the processor configures the next timeout, then transitions to STANDBY mode. Qorivva Simple Cookbook, Rev. 4 120 Freescale Semiconductor 13.3 Code The following includes the main.c code used in the example, plus two preliminary standalone test programs: one for testing only RTC, and the other for RTC with STOP mode. 13.3.1 /* /* /* /* /* main.c: Complete Example (MPC56xxB) main.c - Modes: Low Power example for MPC56xxB/S */ Description: Enters STANDBY, STOP modes and exits at timeout */ Mar 13 2010 S Mihalik - initial version */ Apr 07 2010 S Mihalik - Changed STOP mode config MVRON=0, not 1 */ Copyright Freescale Semiconductor, Inc 2010 All rights reserved. */ #include "MPC5604B_0M27V_0101.h" void disableWatchdog(); void ResetEvent(); void RTCWakeupEvent(); void main (void) { uint16_t ResetCause16 =0U; /* Use proper header file */ /* Prototypes */ /* Placeholder for reset cause status bits */ disableWatchdog(); /* Disable watchdog */ ME.MER.R = 0x0000248D; /* Enable RUN3, STANDBY, STOP, DRUN, other modes */ ME.DRUN.R = 0x001F0010; /* DRUN cfg: 16MHIRCON=1, syclk=16 MHz FIRC */ ME.RUN[3].R = 0x001F0010; /* RUN3 cfg: 16MHIRCON=1, syclk=16 MHz FIRC */ ME.STOP0.R = 0x0015000F; /* 56xxB STOP: FIRCON=0 MVRON=0, flash LP, no sysclk*/ ME.STANDBY0.R = 0x0085000F; /* 56xxB STANDBY cfg: FIRCON = 0 */ ME.RUNPC[7].R = 0x00000088; /* Run Peri. Cfg 7 settings: run in DRUN, RUN3 modes*/ ME.LPPC[7].R = 0x00000400; /* LP Peri. Cfg. 7 settings: run in STOP */ ME.PCTL[68].R = 0x3F; /* MPC56xxB/S SIU: select ME.RUNPC[7], ME.LPPC[7] */ ME.PCTL[69].R = 0x3F; /* MPC56xxB/S WKPU: select ME.RUNPC[7], ME.LPPC[7] */ ME.PCTL[91].R = 0x3F; /* MPC56xxB/S RTC/API: select ME.RUNPC[7], ME.LPPC[7] */ while (ME.GS.B.S_MTRANS) {} ME.MCTL.R = 0x70005AF0; ME.MCTL.R = 0x7000A50F; while (ME.GS.B.S_MTRANS) {} /* Ensure any STANDBY to DRUN mode trans. completed*/ /* Enter RUN3 Mode & Key */ /* Enter RUN3 Mode & Inverted Key */ /* Wait for RUN3 mode transition to complete */ /* Note: could wait here using timer and/or I_TC IRQ */ while(ME.GS.B.S_CURRENTMODE != 7) {} /* Verify RUN3 (0x7) is the current mode */ ResetCause16 = RGM.DES.R; /* Test for destructive reset event: */ if ((uint16_t)ResetCause16 != 0) { /* If destructive reset event */ ResetEvent(); /* and execute ResetEvent function */ } ResetCause16 = RGM.FES.R; /* Test for functional reset event: */ if ((uint16_t)ResetCause16 != 0) { /* If functional reset event */ ResetEvent(); /* and execute ResetEvent function */ } if (WKUP.WISR.R && 0x00000002) { /* MPC56xxB: If wake up event RTC causing reset*/ RTCWakeupEvent(); /* execute RTCWakeupEvent function */ } if (WKUP.WISR.R && 0x00000040) { /*MPC56xxB: If wake up event PE[6] causing reset*/ while(1); /* wait forever */ } } while (1) { } /* else wait forever */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 121 void ResetEvent(void) { CGM.SIRC_CTL.R = 0x00000301; RTC.RTCC.R = 0x00000000; RTC.RTCC.R = 0xA01B1000; WKUP.WIREER.R = 0x00000042; WKUP.WIFER.R = 0x00000040; WKUP.WRER.R = 0x00000042; WKUP.WIPUER.R = 0x000FFFFF; WKUP.WISR.R = 0x000FFFFF; } /* /* /* /* /* /* /* /* MPC56xxB: Div. SIRC by 4, turn SIRC on in STANDBY*/ Clear CNTEN to reset RTC */ CLKSEL=SIRC div. by 4, FRZEN=CNTEN=1, RTCVAL=27 */ MPC56xxB: Enable rising edge events on RTC, PE[0]*/ MPC56xxB: Enable analog filters - , PE[0] */ MPC56xxB: Enable wakeup events for RTC, PE[0] */ MPC56xxB: Enable WKUP pins to stop leakage*/ MPC56xxB: Clear all wake up flags */ RGM.DES.R = 0xFFFF; RGM.FES.R = 0xFFFF; /* Clear destructive reset flags */ /* & clear functional reset flags */ ME.MCTL.R = 0xD0005AF0; ME.MCTL.R = 0xD000A50F; while (1) {} /* Enter STANDBY mode and key */ /* Enter STANDBY mode and inverted key */ /* Wait to enter STANDBY mode (could use timeout) */ /* ENTER STANDBY MODE. ON STANDBY EXIT, RESET VECTOR IS TAKEN IN THIS EXAMPLE */ void RTCWakeupEvent(void) { SIU.PCR[68].R = 0x0200; SIU.PCR[64].R = 0x0103; SIU.GPDO[68].R = 0; RTC.RTCC.R = 0x00000000; RTC.RTCC.R = 0xA0031000; WKUP.WISR.R = 0x00000002; ME.MCTL.R = 0xA0005AF0; ME.MCTL.R = 0xA000A50F; while (ME.GS.B.S_MTRANS) {} /* MPC56xxB/S EVB LED: enable as output */ /* MPC56xxB: Cfg PE[0]input- KEY 1 on MPC56xxB EVB*/ /* MPC56xxB/S EVB LED: data output: LED on */ /* Clear CNTEN to reset RTC, enable reloading RTCVAL*/ /* CLKSEL=SIRC div. by 4, FRZEN=CNTEN=1, RTCVAL=3 */ /* MPC56xxB: Clear wake up flag RTC */ /* Enter STOP mode and key */ /* Enter STOP mode and inverted key */ /* Wait for STOP mode transition to complete */ /* STOP HERE FOR STOP MODE! ON STOP MODE EXIT, CODE CONTINUES HERE:*/ while(ME.GS.B.S_CURRENTMODE != 7) {} /* Verify RUN3 (0x7) is still current mode */ RTC.RTCC.R = 0x00000000; /* Clear CNTEN to reset RTC, enable reloading RTCVAL*/ RTC.RTCC.R = 0xA01B1000; /* CLKSEL=SIRC div. by 4, FRZEN=CNTEN=1, RTCVAL=27 */ WKUP.WISR.R = 0x00000002; /* MPC56xxB: Clear wake up flag RTC */ ME.MCTL.R = 0xD0005AF0; ME.MCTL.R = 0xD000A50F; while (1) {} } /* Enter STANDBY mode and key */ /* Enter STANDBY mode and inverted key */ /* Wait to enter STANDBY mode (could use timeout) */ /* ENTER STANDBY MODE. ON STANDBY EXIT, RESET VECTOR IS TAKEN IN THIS EXAMPLE */ void disableWatchdog(void) { SWT.SR.R = 0x0000c520; /* Write keys to clear soft lock bit */ SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; /* Clear watchdog enable (WEN) */ } Qorivva Simple Cookbook, Rev. 4 122 Freescale Semiconductor 13.3.2 /* /* /* /* /* Test Program 1: Exercise RTC Function only (MPC56xxB) main.c - Modes: Low Power example for MPC56xxB/S - RTC FUNCTION ONLY */ Description: Toggles output based on RTC timeouts */ NOTE: RTCCNT only clears on power up reset, not debugger reset or RESET input */ Mar 12 2010 S Mihalik - Initial version */ Copyright Freescale Semiconductor, Inc 2010 All rights reserved. */ #include "MPC5604B_0M27V_0101.h" /* Use proper header file */ void disableWatchdog(); /* Prototypes */ uint32_t wakeupCtr = 0; /* RTC timeouts */ void main (void) { disableWatchdog(); /* Disable watchdog */ ME.MER.R = 0x0000248D; /* Enable RUN3, STANDBY, STOP, DRUN, other modes */ ME.DRUN.R = 0x001F0010; /* DRUN cfg: 16MHIRCON=1, syclk=16 MHz FIRC */ ME.RUN[3].R = 0x001F0010; /* RUN3 cfg: 16MHIRCON, syclk=16 MHz FIRC */ ME.RUNPC[7].R = 0x00000088; /* Run Peri Cfg 7 settings: run in DRUN, RUN3 modes */ ME.PCTL[68].R = 0x3F; /* MPC56xxB/S SIU: select ME.RUNPC[7], ME.LPPC[7] */ ME.PCTL[91].R = 0x3F; /* MPC56xxB/S RTC/API: select ME.RUNPC[7], ME.LPPC[7]*/ /* Mode Transition to enter RUN3 mode: */ ME.MCTL.R = 0x70005AF0; /* Enter RUN3 Mode & Key */ ME.MCTL.R = 0x7000A50F; /* Enter RUN3 Mode & Inverted Key */ while (ME.GS.B.S_MTRANS) {} /* Wait for RUN3 mode transition to complete */ /* Note: could wait here using timer and/or I_TC IRQ*/ while(ME.GS.B.S_CURRENTMODE != 7) {} /* Verify RUN3 (0x7) is the current mode */ } SIU.GPDO[68].R = 1; /* SIU.PCR[68].R = 0x0200; /* CGM.SIRC_CTL.R = 0x00000301;/* RTC.RTCC.R = 0x00000000; /* RTC.RTCC.R = 0xA01B1000; /* MPC56xxB/S EVB LED: data output: LED off */ MPC56xxB/S EVB LED: enable as outputt */ MPC56xxB: Div. SIRC by 4 & turn SIRC on in STANDBY*/ Clear CNTEN to reset RTC (counter) */ CLKSEL=SIRC (div. by 4), FRZEN=CNTEN=1, RTCVAL=27*/ while (1) { while (RTC.RTCS.B.RTCF == wakeupCtr++; SIU.GPDO[68].R = 0; RTC.RTCC.R = 0x00000000; RTC.RTCC.R = 0xA0031000; RTC.RTCS.R = 0x20000000; 0) /* /* /* /* /* {} /* Wait for RTC timeout */ Increment wakeup counter */ MPC56xxB/S EVB LED: data output: LED on */ Clear CNTEN to reset RTC, enable reloading RTCVAL*/ CLKSEL=SIRC div. by 4, FRZEN=CNTEN=1, RTCVAL=3 */ Clear RTC flag */ while (RTC.RTCS.B.RTCF == SIU.GPDO[68].R = 1; RTC.RTCC.R = 0x00000000; RTC.RTCC.R = 0xA01B1000; RTC.RTCS.R = 0x20000000; 0) /* /* /* /* {} /* Wait for RTC timeout */ MPC56xxB/S EVB LED: data output: LED off */ Clear CNTEN to reset RTC, enable reloading RTCVAL*/ CLKSEL=SIRC div. by 4, FRZEN=CNTEN=1, RTCVAL=27*/ Clear RTC flag */ } void disableWatchdog(void) { SWT.SR.R = 0x0000c520; /* Write keys to clear soft lock bit */ SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; /* Clear watchdog enable (WEN) */ } Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 123 13.3.3 /* /* /* /* /* Test Program 2: Exercise RTC Wakeup in STOP mode (MPC56xxB) main.c - Modes: Low Power example for MPC56xxB/S - RTC, WAKEUP, STOP MODE ONLY */ Description: Toggles output on RTC timeouts */ NOTE: RTCCNT only clears on power up reset, not debugger reset or RESET input */ Mar 13 2010 S Mihalik - Initial version */ Copyright Freescale Semiconductor, Inc 2010 All rights reserved. */ #include "MPC5604B_0M27V_0101.h" void disableWatchdog(); uint32_t wakeupCtr = 0; uint32_t temp = 0; /* Use proper header file */ /* Prototypes */ /* wake up counter incremented on STANDBY exit */ void main (void) { disableWatchdog(); /* Disable watchdog */ ME.MER.R = 0x0000248D; /* Enable RUN3, STANDBY, STOP, DRUN, other modes */ ME.DRUN.R = 0x001F0010; /* DRUN cfg: 16MHIRCON=1, syclk=16 MHz FIRC */ ME.RUN[3].R = 0x001F0010; /* RUN3 cfg: 16MHIRCON, syclk=16 MHz FIRC */ ME.STOP0.R = 0x0015001F; /* 56xxB STOP cfg: FIRCON=MVRON=1, flashLP, no sysclk*/ ME.RUNPC[7].R = 0x00000088; /* Run Peri. Cfg 7 settings: run in DRUN, RUN3 modes*/ ME.LPPC[7].R = 0x00000400; /* LP Peri. Cfg 7 settings: run in STOP */ ME.PCTL[68].R = 0x3F; /* MPC56xxB/S SIU: select ME.RUNPC[7], ME>LPPC[7] */ ME.PCTL[69].R = 0x3F; /* MPC56xxB/S WKPU: select ME.RUNPC[7], ME>LPPC[7] */ ME.PCTL[91].R = 0x3F; /* MPC56xxB/S RTC/API: select ME.RUNPC[7], ME.LPPC[7]*/ ME.MCTL.R = 0x70005AF0; /* Enter RUN3 Mode & Key */ ME.MCTL.R = 0x7000A50F; /* Enter RUN3 Mode & Inverted Key */ while (ME.GS.B.S_MTRANS) {} /* Wait for RUN3 mode transition to complete */ /* Note: could wait here using timer and/or I_TC IRQ*/ while(ME.GS.B.S_CURRENTMODE != 7) {} /* Verify RUN3 (0x7) is the current mode */ CGM.SIRC_CTL.R = 0x00000301; RTC.RTCC.R = 0x00000000; RTC.RTCC.R = 0xA01B1000; WKUP.WIREER.R = 0x00000002; WKUP.WIFER.R = 0x00000000; WKUP.WRER.R = 0x00000002; WKUP.WIPUER.R = 0x000FFFFF; /* /* /* /* /* /* /* MPC56xxB: Div SIRC by 4 & turn SIRC on in STANDBY*/ Clear CNTEN to reset RTC (counter) */ CLKSEL=SIRC div. by 4, FRZEN=CNTEN=1, RTCVAL=27 */ MPC56xxB: Enable rising edge events on RTC */ MPC56xxB: Enable analog filters - none */ MPC56xxB: Enable wakeup events for RTC */ MPC56xxB: Ena. WKUP pins pullups to stop leakage*/ ME.MCTL.R = 0xA0005AF0; ME.MCTL.R = 0xA000A50F; while (ME.GS.B.S_MTRANS) {} /* Enter STOP mode and key */ /* Enter STOP mode and inverted key */ /* Wait STOP mode transition to complete */ /* STOP HERE FOR STOP MODE! ON STOP MODE EXIT, CODE CONTINUES HERE:*/ while (1) { while(ME.GS.B.S_CURRENTMODE != 7) {} /* Verify RUN3 (0x7) is the current mode */ wakeupCtr++; /* Increment wakeup counter */ SIU.GPDO[68].R = 0; SIU.PCR[68].R = 0x0200; /* MPC56xxB/S EVB LED: data output: LED on */ /* MPC56xxB/S EVB LED: enable as output */ RTC.RTCC.R = 0x00000000; /* Clear CNTEN to reset RTC, enable reloading RTCVAL*/ RTC.RTCC.R = 0xA0031000; /* CLKSEL=SIRC div. by 4, FRZEN=CNTEN=1, RTCVAL=3 */ WKUP.WISR.R = 0x00000002; /* MPC56xxB: Clear wake up flag RTC */ ME.MCTL.R = 0xA0005AF0; /* Enter STOP mode and key */ ME.MCTL.R = 0xA000A50F; /* Enter STOP mode and inverted key */ while (ME.GS.B.S_MTRANS) {} /* Wait STOP mode transition to complete */ /* STOP HERE FOR STOP MODE! ON STOP MODE EXIT, CODE CONTINUES HERE:*/ Qorivva Simple Cookbook, Rev. 4 124 Freescale Semiconductor while(ME.GS.B.S_CURRENTMODE != 7) {} /* Verify RUN3 (0x7) is the current mode */ SIU.GPDO[68].R = 1; RTC.RTCC.R = 0x00000000; RTC.RTCC.R = 0xA01B1000; WKUP.WISR.R = 0x00000002; /* /* /* /* MPC56xxB/S EVB LED: data output: LED off */ Clear CNTEN to reset RTC, enable reloading RTCVAL*/ CLKSEL=SIRC (div. by 4), FRZEN=CNTEN=1, RTCVAL=27*/ MPC56xxB: Clear wake up flag RTC */ ME.MCTL.R = 0xA0005AF0; /* Enter STOP mode and key */ ME.MCTL.R = 0xA000A50F; /* Enter STOP mode and inverted key */ while (ME.GS.B.S_MTRANS) {} /* Wait STOP mode transition to complete */ } } /* STOP HERE FOR STOP MODE! ON STOP MODE EXIT, CODE CONTINUES HERE:*/ void disableWatchdog(void) { SWT.SR.R = 0x0000c520; SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; } /* Write keys to clear soft lock bit */ /* Clear watchdog enable (WEN) */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 125 14 eDMA: Block Move 14.1 Description Task: Initialize an eDMA channel’s Transfer Control Descriptor (TCD) to transfer a string of bytes (“Hello world”) from an array in SRAM to a single SRAM byte location. This emulates a common use of DMA, where a string of data or commands is transferred automatically under DMA control to an input register of a peripheral. Only one byte of data will be transferred with each DMA service request. Hence the “minor loop” is simply one transfer, which transfers one byte. The “major loop” here consists of 12 minor loop iterations. Because a peripheral is not involved, automatic DMA handshaking will not occur. Instead, the software handshaking given here must be implemented for each transfer: • Start DMA service request (set a START bit). • Poll when that request is done (check the CITER bit field). These steps appear “messy” for every transfer, which is only a byte in this example. However, remember that when using actual peripherals, software never has to do these steps; they are done automatically by hardware. The purpose of this example is to illustrate how to set up a DMA transfer. TCD0 will be used. On MPC555x devices, this TCD is assigned to the eQADC’s Command FIFO 0. On MPC551x devices, TCDs are not hard-assigned to any channel. Because TCDs are in RAM, all fields will come up random. Hence all fields used or enabled must be initialized. Exercise: Step through code, observing DMA transferring the “Hello world” string to the destination byte. Then modify the TCD so the destination is an array instead of a single byte location. MPC5500 eDMA SRAM Transfer Control Descriptors (TCD’s) SRAM TCD0 TCD1 TCD2 TCD3 TCD4 •• • TCDn Source Array XBAR Destination Byte Other XBAR Masters Other XBAR Slaves Figure 25. DMA Block Move Example Qorivva Simple Cookbook, Rev. 4 126 Freescale Semiconductor 14.2 Design Table 57. DMA: Block Move Step Relevant Bit Fields Data Init Initialize data source string. Init TCD0 Source Parameters: • Source address: address of SourceData • Source data transfer size: read 1 byte per transfer • Source address offset: increment source address by 1 after each transfer • Last source address adjustment: adjust source address by –12 at completion of major loop • Source address modulo: disabled SourceData = “Hello world “ Destination Parameters: • Destination address: address of SourceData • Destination data transfer size: write one byte per transfer • Destination address offset: do not increment destination address after each transfer • Last destination address adjustment: do not adjust destination address after completion of major loop • Destination address modulo: disabled SADDR = src. addr. SSIZE = 0 (1 byte) SOFF = 1 SLAST = –12 SMOD = 0 DADDR = dest. addr. DSIZE = 0 (1 byte) Initiate Set channel 0 START bit to initiate first minor loop DMA For each transfer (while not on last transfer)1: service by • Wait for last transfer to have started and minor loop software is finished (is no longer active) • Set start bit to initiate next minor loop transfer 1 TCD0[SLAST] = –12 TCD0[SMOD] = 0 TCD0[DADDR] = &Dest. TCD0[DSIZE] = 0 TCD0[DOFF] = 0 DLAST = 0 TCD0[DLAST_SGA] = 0 DMOD = 0 TCD0[DMOD] = 0 Init DMA Use fixed priorities for groups and channels (default) Arbitration Use defaults: Priorities (Gp = 0, Ch = 0), no preemption Set enable request for channel 0 TCD0[SADDR] = &SourceData[] TCD0[SSIZE] = 0 TCD0[SOFF] = 1 DOFF = 0 Control Parameters • Inner “minor loop” byte transfer count: transfer 1 byte NBYTES = 1 per service request • Number of minor loop iterations in major loop: 12 BITER = 12 CITER = 12 • Disable request: at end of major loop, disable further D_REQ = 1 requests for transfer • Interrupts: interrupts are not enabled here INT_HALF = 0 INT_MAJOR = 0 • Linking: channel linking is not enabled here CITERE_LINK = 0 BITERE_LINK = 0 • Dynamic programming: dynamic channel linking MAJORE_LINK = 0 and scatter gather are not enabled E_SG = 0 • Bandwidth Control: no DMA stalls BWC = 0 • Status flags: initialize to 0 START = 0 DONE = 0 ACTIVE = 0 Enable Chan. 0 Pseudo Code TCD0[NBYTES] = 1 TCD0[BITER] = 12 TCD0[CITER] = 12 TCD0[D_REQ] = 1 TCD0[INT_HALF] = 0 TCD0[INT_MAJ] = 0 TCD0[CITERE_LINK] = 0 TCD0[BITERE_LINK] = 0 TCD0[MAJORE_LINK] = 0 TCD0[E_SG] = 0 TCD0[.BWC] = 0 TCD0[START] = 0 TCD0[DONE] = 0 TCD0[ACTIVE] = 0 EDMA_CR = 0x0000 E400 EDMA_CPR0 = 0x00 SERQ = 0 SSB = 0 wait (START = 1 and ACTIVE = 0) SERQ = 0 EDMA_SERQR = 0 EDMA_SSBR = 0 while (TCD0[CITER] != 1) { wait for ((TCD0[START] = 1) & (TCD0[ACTIVE] = 0)) EDMA_SSBR = 0 } The START bit is set on the last iteration here, but is not used. After completion, the channel’s DONE bit will be set. Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 127 14.3 Code 14.3.1 /* /* /* /* /* /* main.c (All except MPC56xxS) main.c - DMA-Block Move example */ Rev 0.1 Sept 30, 2004 S.Mihalik, Copyright Freescale, 2004. All Rights Reserved */ Rev 1.0 Jul 10 2007 SM - Changed from TCD18 to TCD0 to be MPC5510 compatible */ Notes: */ 1. MMU not initialized; must be done by debug scripts or BAM */ 2. L2SRAM not initialized; must be done by debug scripts */ #include "mpc563m.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ const uint8_t uint8_t SourceData[] = {"Hello World\r"}; /* Source data string */ Destination = 0; /* Destination byte */ void initTCD0(void) { EDMA.TCD[0].SADDR = (vuint32_t) &SourceData; EDMA.TCD[0].SSIZE = 0; EDMA.TCD[0].SOFF = 1; EDMA.TCD[0].SLAST = -12; EDMA.TCD[0].SMOD = 0; /* /* /* /* /* Load address of source data */ Read 2**0 = 1 byte per transfer */ After transfer, add 1 to src addr*/ After major loop, reset src addr*/ Source modulo feature not used */ EDMA.TCD[0].DADDR = (vuint32_t) &Destination; /* Load address of destination */ EDMA.TCD[0].DSIZE = 0; /* Write 2**0 = 1 byte per transfer */ EDMA.TCD[0].DOFF = 0; /* Do not increment destination addr*/ EDMA.TCD[0].DLAST_SGA = 0; /* After major loop, no dest addr change*/ EDMA.TCD[0].DMOD = 0; /* Destination modulo feature not used */ } EDMA.TCD[0].NBYTES = 1; EDMA.TCD[0].BITER = 12; EDMA.TCD[0].CITER = 12; EDMA.TCD[0].D_REQ = 1; EDMA.TCD[0].INT_HALF = 0; EDMA.TCD[0].INT_MAJ = 0; EDMA.TCD[0].CITERE_LINK = 0; EDMA.TCD[0].BITERE_LINK = 0; EDMA.TCD[0].MAJORE_LINK = 0; EDMA.TCD[0].E_SG = 0; EDMA.TCD[0].BWC = 0; EDMA.TCD[0].START = 0; EDMA.TCD[0].DONE = 0; EDMA.TCD[0].ACTIVE = 0; void main (void) { int i = 0; initTCD0(); /* /* /* /* /* Transfer 1 byte per minor loop */ 12 minor loop iterations */ Initialize current iteraction count */ Disable channel when major loop is done*/ Interrupts are not used */ /* Linking is not used */ /* Dynamic program is not used */ /* Default bandwidth control- no stalls */ /* Initialize status flags */ /* Dummy idle counter */ /* Initialize DMA Transfer Control Descriptor 0 */ EDMA.CR.R = 0x0000E400; /* Use fixed priority arbitration for DMA groups and channels */ EDMA.CPR[0].R = 0x12; /* Channel 0 priorites: group priority = 1, channel priority = 2 */ EDMA.SERQR.R = 0; /* Enable EDMA channel 0 */ EDMA.SSBR.R = 0; /* Set channel 0 START bit to initiate first minor loop transfer */ /* Initate DMA service using software */ while (EDMA.TCD[0].CITER != 1) { /* while not on last minor loop */ /* wait for START=0 and ACTIVE=0 */ while ((EDMA.TCD[0].START == 1) | (EDMA.TCD[0].ACTIVE == 1)) {} EDMA.SSBR.R = 0; /* Set channel 0 START bit again for next minor loop transfer */ } } while (1) { i++; } /* Loop forever */ Qorivva Simple Cookbook, Rev. 4 128 Freescale Semiconductor 14.3.2 /* /* /* /* main.c (MPC56xxS) main.c - DMA Block Move */ Oct 23 2008 S.Mihalik -inital version */ Mar 15 2010 S Mihalik - updated for new eDMA version on MPC5606S */ Copyright Freescale Semiconductor, Inc 2008, 2010 All rights reserved. */ #include "56xxS_0204.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ const uint8_t SourceData[] = {"Hello World\r"};/* Source data string */ uint8_t Destination = 0; /* Destination byte */ void initTCD0(void) { EDMA.TCD[0].SADDR = (vuint32_t) &SourceData; EDMA.TCD[0].SSIZE = 0; EDMA.TCD[0].SOFF = 1; EDMA.TCD[0].SLAST = -12; EDMA.TCD[0].SMOD = 0; /* /* /* /* /* Load address of source data */ Read 2**0 = 1 byte per transfer */ After transfer, add 1 to src addr*/ After major loop, reset src addr*/ Source modulo feature not used */ EDMA.TCD[0].DADDR = (vuint32_t) &Destination; /* Load address of destination */ EDMA.TCD[0].DSIZE = 0; /* Write 2**0 = 1 byte per transfer */ EDMA.TCD[0].DOFF = 0; /* Do not increment destination addr */ EDMA.TCD[0].DLAST_SGA = 0; /* After major loop, no dest addr change*/ EDMA.TCD[0].DMOD = 0; /* Destination modulo feature not used */ } EDMA.TCD[0].NBYTESu.R = 1; EDMA.TCD[0].BITER = 12; EDMA.TCD[0].CITER = 12; EDMA.TCD[0].D_REQ = 1; EDMA.TCD[0].INT_HALF = 0; EDMA.TCD[0].INT_MAJ = 0; EDMA.TCD[0].CITERE_LINK = 0; EDMA.TCD[0].BITERE_LINK = 0; EDMA.TCD[0].MAJORE_LINK = 0; EDMA.TCD[0].E_SG = 0; EDMA.TCD[0].BWC = 0; EDMA.TCD[0].START = 0; EDMA.TCD[0].DONE = 0; EDMA.TCD[0].ACTIVE = 0; void main (void) { volatile uint32_t i = 0; initTCD0(); /* /* /* /* /* Transfer 1 byte per minor loop */ 12 minor loop iterations */ Initialize current iteraction count */ Disable channel when major loop is done*/ Interrupts are not used */ /* Linking is not used */ /* Dynamic program is not used */ /* Default bandwidth control- no stalls */ /* Initialize status flags */ /* Dummy idle counter */ /* Initialize DMA Transfer Control Descriptor 0 */ EDMA.CR.R = 0x0000E400; /* Use fixed priority arbitration for DMA groups and channels */ EDMA.CPR[0].R = 0x0; /* Channel 0 priorites: group priority = 0, channel priority = 0 */ EDMA.SERQ.R = 0; EDMA.SSRT.R = 0; } /* Enable EDMA channel 0 */ /* Set channel 0 START bit to initiate first minor loop transfer */ /* Initate DMA service using software */ while (EDMA.TCD[0].CITER != 1) { /* while not on last minor loop */ /* wait for START=0 and ACTIVE=0 */ while ((EDMA.TCD[0].START == 1) | (EDMA.TCD[0].ACTIVE == 1)) {} EDMA.SSRT.R = 0; /* Set channel 0 START bit to initiate first minor loop transfer */ } while (1) { LoopForever: i++; } /* Loop forever */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 129 15 eSCI: Simple Transmit and Receive 15.1 Description Task: Transmit the string of data “Hello World<CR>” to eSCI_A, then wait to receive a byte. The serial parameters are: 9600 baud, eight bits data, and no parity. Interrupts and DMA are not used. The program simply waits until flags are at the desired state before proceeding. Overflow is not checked in this example, as we look for only one received character. If overflow is a concern on reception, the overflow flag would be checked after reading the received character. Exercise: Connect a COM port of a PC to an MPC5500 evaluation board eSCI port. Use a terminal emulation program for communication from the PC side with settings of 9600 baud, 8 bits data, no parity, and no flow control. Make sure to have the correct SCI jumper settings for Tx/Rx connection for the particular EVB used. Step through the program and verify proper transmission. MPC5500 ‘Hello World<CR>’ eSCI_A 8 MHz Crystal PLL Set to 64 MHz Data Register TxDA Personal Computer COM Port XCVR RxDA 64 MHz sysclk Received Characters Figure 26. eSCI Simple Transmit and Receive Example Table 58. Signals for eSCI Example MPC551x Family Signal Pin Name SIU PCR No. TxDA PD6 RxDA PD7 MPC555x Family Package Pin No. Function Name SIU PCR No. 144 QFP 176 QFP 208 BGA 54 98 122 F14 TXDA 55 97 121 F15 RXDA Package Pin No. EVB 496 BGA 416 BGA 324 BGA 208 BGA xPC 563M 89 V24 U24 N20 J14 PJ1–14 90 U26 V24 P20 K14 PJ1–13 Qorivva Simple Cookbook, Rev. 4 130 Freescale Semiconductor 15.2 Design For the desired 9600 SCI baud rate, the ideal calculation for SBR bitfield eSCIx_CR1[SBR], assuming a 64 MHz sysclk to the eSCI module, is: eSCI system clock 64 M SBR = --------------------------- = ------------- = 416.66... 16 SCI Baud Rate 16 9600 So after rounding it off to the nearest whole number, 417 (0x1A1) will be used for SBR in this example. The time per bit at 9600 baud = 1 sec / 9600 bits, which rounds off to 104 microseconds. The oscilloscope shot below shows only one frame (character) because it was taken when stepping through code. It shows the bit time being met. The scope shot also shows the transmission sequence of sending the start bit, followed by LSB through MSB bits. Here the data bits, LSB first, are 0001 0010, which after putting MSB first is 0x48, the ASCII code for “H.” The scope shot below shows the program running without a breakpoint. There are contiguous transmissions of characters, and the first two character are shown here. Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 131 Table 59. eSCI Simple Design Steps Relevant Bit Fields Step Pseudo Code MPC551x MPC555x Data Init Initialize data string to be transmitted on eSCI_A Initialize byte used to receive data from eSCI_A TransData = “Hello World!<CR>” RecData = 0 Init sysclk Set system clock = 64 MHz (See Section 10, “PLL: Initializing System Clock (MPC551x, MPC55xx)”) CRP_CLKSRC FMPLL_SYNCR = [XOSCEN] = 1; 0x1608 0000; PLL_ESYNCR2 = Wait for 0x0000 0006; FMPLL_SYNSR PLL_ESYNCR1 = [LOCK] = 1; 0xF000 0020; FMPLL_SYNCR = Wait for 0x1600 0000; PLL_SYNSR [LOCK] = 1; Also for PLL_ESYNCR2 = MPC563x: 0x0000 0005; FMPLL_ SIU_SYSCLK ESYNCR1 [SYSCLKSEL] = 2; [CLKCFG]=7 Note for 40 MHz Crystal used on MPC55xx — Replace FMPLL_SYNCR values: 0x1608 0000 with 0x4610 0000 0x1600 0000 with 0x4608 0000 Init eSCI_A Turn on the eSCI module (in case it was off) MDIS = 0 (default) Initialize eSCI control • Baud Rate value = 64 M / (16 9600) ~= 417 • Word length = 8 bits • Parity is not enabled • Enable transmitter • Enable receiver SBR = 417 (0x1A1) M = 0 (default) PE = 0 (default) TE = 1 RE = 1 Configure pads: • TxDA PA = primary • RxDA PA = primary Transmit Data Loop for # characters: • Wait for Transmit Data Register Empty status flag • Clear status flag • Write one character Read and Echo Back Data ESCIA_CR2 = 0x2000 ESCIA_CR1 = 0x01A1 000C SIU_PCR[54] = 0x0400 SIU_PCR[55] = 0x0400 SIU_PCR[89] = 0x0400 SIU_PCR[90] = 0x0400 wait for TDRE = 1 write TRDE = 1 Loop for # characters { wait for ESCIA_SR[TDRE] = 1 ESCIA_SR = 0x8000 0000 ESCIA_DR = next char. } Wait for Receive Data Register Full status flag wait for RDRF = 1 wait for ESCIA_SR[RDRF] = 1 Clear status flag write RDRF = 1 Read byte ESCIA_SR = 0x2000 0000 RecData = ESCIA_DR Ensure Transmit Data Register Empty status flag wait for TDRE = 1 Clear status flag write TRDE = 1 Transmit back (echo) received byte wait for ESCIA_SR[TDRE] = 1 ESCIA_SR = 0x8000 0000 ESCIA_DR = RecData Qorivva Simple Cookbook, Rev. 4 132 Freescale Semiconductor 15.3 /* /* /* /* /* /* /* /* /* Code main.c: Rev 0.1 Rev 1.0 Rev 1.1 Simple eSCI program */ Sept 30, 2004 S.Mihalik, Copyright Freescale, 2007. All Rights Reserved */ July 14 2005 SM-Cleared TDRE,RDRF flags as required in MPC5554 RevA & on*/ Jul 19 2007 SM - Modified for MPC551x, changed sysclk (50 MHz) & SBR, */ cleared RDRF before reading data register */ Rev 1.2 Aug 08 2007 SM - Changed sysclk to 64 MHz */ Rev 1.3 Jun 04 2008 SM - initSysclk changed for MPC5633M support */ Notes: 1 MMU not initialized; must be done by debug scripts or BAM */ 2 SRAM not initialized; must be done by debug scripts or in a startup file*/ #include "mpc563m.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ const uint8_t uint8_t TransData[] = {"Hello World!\n\r"}; /* Transmit string & CR*/ RecData; /* Received byte from eSCI */ void initSysclk (void) { /* MPC551x: Use the next 6 lines */ /* CRP.CLKSRC.B.XOSCEN = 1; */ /* Enable external oscillator */ /* FMPLL.ESYNCR2.R = 0x00000006; */ /* Set ERFD to initial value of 6 */ /* FMPLL.ESYNCR1.R = 0xF0000020; */ /* Set CLKCFG=PLL, EPREDIV=0, EMFD=0x20*/ /* while (FMPLL.SYNSR.B.LOCK != 1) {};*/ /* Wait for PLL to LOCK */ /* FMPLL.ESYNCR2.R = 0x00000005; */ /* Set ERFD to final value for 64 MHz sysclk /* SIU.SYSCLK.B.SYSCLKSEL = 2; */ /* Select PLL for sysclk */ /* MPC563x: Use the next line */ FMPLL.ESYNCR1.B.CLKCFG = 0X7; /* Change clk to PLL normal mode from crystal /* MPC555x including MPC563x: use the next 3 lines for either 8 or 40 MHz crystal FMPLL.SYNCR.R = 0x16080000; /* 8 MHz xtal: 0x16080000; 40MHz: 0x46100000 while (FMPLL.SYNSR.B.LOCK != 1) {}; /* Wait for FMPLL to LOCK */ FMPLL.SYNCR.R = 0x16000000; /* 8 MHz xtal: 0x16000000; 40MHz: 0x46080000 } void initESCI_A (void) { ESCI_A.CR2.R = 0x2000; ESCI_A.CR1.R = 0x01A1000C; /* Use the following two lines /* SIU.PCR[54].R = 0x400; /* SIU.PCR[55].R = 0x400; /* Use the following two lines SIU.PCR[89].R = 0x400; SIU.PCR[90].R = 0x400; } */ */ */ */ */ /* Module is enabled (default setting ) */ /* 9600 baud, 8 bits, no parity, Tx & Rx enabled */ for MPC551x */ /* Configure pad for primary func: TxDA */ /* Configure pad for primary func: RxDA */ for MPC555x */ /* Configure pad for primary func: TxDA */ /* Configure pad for primary func: RxDA */ void TransmitData (void) { uint8_t j; /* Dummy variable */ for (j=0; j< sizeof (TransData); j++) { /* Loop for character string */ while (ESCI_B.SR.B.TXRDY == 0) {} /* Wait for LIN transmit ready = 1 */ ESCI_B.SR.R = 0x00004000; /* Clear TXRDY flag */ ESCI_B.LTR.R = FrameSpecAndData[j] << 24; /* Write byte to LIN Trans Reg. */ } } void ReceiveData (void) { while (ESCI_A.SR.B.RDRF == 0) {} ESCI_A.SR.R = 0x20000000; RecData = ESCI_A.DR.B.D; while (ESCI_A.SR.B.TDRE == 0) {} ESCI_A.SR.R = 0x80000000; ESCI_A.DR.B.D = RecData; } void main(void) { initSysclk(); initESCI_A(); TransmitData(); ReceiveData(); while (1) {} } /* /* /* /* /* /* /* /* /* /* /* Wait for receive data reg full = 1 */ Clear RDRF flag */ Read byte of Data*/ Wait for transmit data reg empty = 1 */ Clear TDRE flag */ Echo back byte of Data read */ Set sysclk = 64 MHz running from PLL */ Enable Tx/Rx for 9600 baud, 8 bits, no parity */ Transmit string of characters on eSCI A */ Receive and echo character on eSCI A */ Wait forever */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 133 16 eSCI: LIN Transmit 16.1 Description Task: Transmit the data string “Hello” via LIN using eSCI_B. When sending bytes to the LIN Transmit Register, use software to poll the TXRDY flag for each byte write, with software clearing the flag each time. For the most efficiency, DMA would be used, which automatically waits for TXRDY flag and clears it after a write. The next most efficient method would be to have the TXRDY interrupt the processor. Parameters from Section 15, “eSCI: Simple Transmit and Receive,” are used, including 10417 baud, 8 bits data, and no parity. Interrupts and DMA are not used. NOTE The RxD receive pin must be able to monitor the TxD transmitted signal. Hence either (1) the LIN transceiver must be connected as shown and the transceiver must be powered, or (2) for debug purposes, the TxD and RxD can be shorted together without using the transceiver. The MPC5510 EVB and MPC563M EVB require changing the default configuration to implement one of these two options. Exercise: Connect an oscilloscope or LIN tool and verify the transmit output. Alter frame ID and repeat. MPC5500 LIN frame specification and data eSCI_B 8 MHz Crystal PLL Set to 64 MHz LIN Transmit Register TxDB LIN XCVR LIN Receive Register RxDB Scope or LIN tool 64 MHz sysclk Figure 27. eSCI LIN Example Table 60. Signals for eSCI LIN Example MPC551x Family Signal Pin Name SIU PCR No. TxDB PD8 RxDB PD9 MPC555x Family Package Pin No. Function Name SIU PCR No. 144 QFP 176 QFP 208 BGA 56 94 118 G13 TXDB 57 93 117 F16 RXDB Package Pin No. EVB 496 BGA 416 BGA 324 BGA 208 BGA xPC 563M 91 Y27 W25 R21 L13 PJ1–16 92 Y24 W23 T19 M13 PJ1–15 Qorivva Simple Cookbook, Rev. 4 134 Freescale Semiconductor 16.2 Design To transmit a LIN frame, first the frames data must be specified by writing the ID, length and control to the LIN Transmit Register. This specifies the LIN frame, and the frame’s header starts to transfer. Next, the data is written to the same LIN Transmit Register. The byte values used in the example are shown in the table below. Per LIN 2.x and J2602 the a checksum is used (CSUM = 1), which includes the header (HDCHK = 1). Table 61. Bytes Written to LIN Transmit Register (SCIB_LTR) Field Parity (1:0) = 0 (Parity not used) Byte Value ID (5:0) = 0x35 Length (7:0) = 8 (for 8 bytes data) HDCHK = 1 (header is included in checksum) CSUM = 1 (checksum is appended to end of frame) CRC = 0 (2 CRC bytes not appended to end of frame) TX = 1 (Transmit operation) 0x35 0x08 Timeout (11:8) = 0 (Timeout is zero for transmit frame) Notes 0xD0 Data 0 = ‘H’ 0x48 Data 1 = ‘e’ 0x65 Data 2 = ‘l’ 0x6C Data 3 = ‘l’ 0x6C Data 4 = ‘o’ 0x6F Data 5 = ‘ ’ 0x20 Data 6 = ‘ ’ 0x20 Data 7 = ‘ ’ 0x20 Specifies LIN frame (which causes LIN frame’s header to start transmission) Data to transmit in LIN frame Bit or physical bus errors can stop DMA transmission (by setting the BSTP bit in eSCIx_CR2). Although DMA is not used here, this bit will be set anyway. This example shows a “good” case, where no errors occur or are checked. Normally one would enable error interrupts and/or check those status flags also. A partial list of errors includes: • A physical bus error of a permanently low bit: sets framing error flag, ESCIx_SR[FE] • RxD input stuck after transmission starts: sets physical bus error flag, ESCIx_LSR[PBERR] • For receive frames, a slave does not respond in the specified timeout in ESCIx_LTR: sets slave timeout flag, ESCIx_LSR_[STO] The LIN state machine can automatically reset after an exception of a bit error, physical bus error, or wakeup. For debug purposes, it is useful to disable this feature, but we will allow the reset to occur by disabling the feature (accomplished by clearing the LDBG bit in eSCIx_LCR). Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 135 16.2.1 Design Steps Table 62. eSCI LIN Design Steps Relevant Bit Fields Step Data Init LIN frame specification and 8 bytes data for transmit Init sysclk Set system clock = 64 MHz (See Section 10, “PLL: Initializing System Clock (MPC551x, MPC55xx)”) Transmit Data ESCB_CR2 = 0x6240 MDIS = 0 (default) BRK13 = 1 BSTP = 1 SBSTP = 1 FBR = 1 ESCIB_CR1 = 0x0180 000C SBR = 384 (0x180) M = 0 (default) PE = 0 (default) TE = 1 RE = 1 TIE = TCIE = RIE = 0 Initialize desired LIN Control Register settings: • Switch eSCI to LIN mode • Do not enable exception to reset state machine LIN = 1 LDBG = 0 Configure pad: • TxDB PA = primary Loop for each FrameSpecAndData character: • Wait for Transmit Data Ready status flag • Clear status flag • Write one character MPC555x CRP_CLKSRC FMPLL_SYNCR = [XOSCEN] = 1; 0x1608 0000; PLL_ESYNCR2 = Wait for 0x0000 0006; FMPLL_SYNSR PLL_ESYNCR1 = [LOCK] = 1; 0xF000 0020; FMPLL_SYNCR = Wait for 0x1600 0000; PLL_SYNSR [LOCK] = 1; Also for PLL_ESYNCR2 = MPC563x: 0x0000 0005; FMPLL_ SIU_SYSCLK ESYNCR1 [SYSCLKSEL] = 2; [CLKCFG] = 7 Initialize desired Control Register 1 settings: • Baud Rate value = 64 M / (16x10417) ~= 384 • Word length = 8 bits • Parity is not enabled • Enable transmitter • Enable receiver • Do not use normal eSCI interrupts • RxDB (NOTE: Some EVB transceivers, such as TJA1020, require a pullup. In this case, at least the internal pullup should be used, so the PCR value should be 0x0403 instead of 0x400.) MPC551x FrameSpecAndData = 0x35, 0x08, 0xD0, ‘H’, ‘e’, ‘l’, ‘l’, ‘o’, ‘ ’, ‘ ’, ‘ ’ Note for 40 MHz Crystal used on MPC55xx — Replace FMPLL_SYNCR values: 0x1608 0000 with 0x4610 0000 0x1600 0000 with 0x4608 0000 Init eSCI B Initialize desired Control Register 2 settings: for LIN • Turn on the eSCI module (in case it was off) • Set break character length to 13 bits • Stop DMA (not used here) on bit/physical bus error • SCI bit errors cause stop on bit instead of frame end • Detect bit errors on each bit (“fast”) instead of byte Pseudo Code ESCIB_LCR = 0x0100 0000 PA = primary wait for TXRDY = 1 write TXRDY = 1 write byte to LTR SIU_PCR[56] = 0x0400 SIU_PCR[57] = 0x0400 SIU_PCR[91] = 0x0400 SIU_PCR[92] = 0x0400 Loop for # characters { wait for ESCIB_SR[TXRDY] = 1 ESCIB_SR = 0x0000 4000 ESCIB_LTR = next char. } Qorivva Simple Cookbook, Rev. 4 136 Freescale Semiconductor 16.2.2 Design Screenshots Below is an oscilloscope trace of TxDB output showing the frame’s header. Normally the data would continue after the header, but code was deliberately stepped through here to show the header generation after the first three writes to the LIN Transmit Register. Figure 28. LIN Transmit Frame Header Figure 29. Entire LIN Transmit Frame Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 137 16.3 /* /* /* /* /* /* /* Code main.c: Simple eSCI LIN program */ Rev 0.1 Oct 10, 2007 S.Mihalik- Initial version */ Rev 0.2 Jun 04 2008 SM - initSysclk changed for MPC5633M support */ Rev 0.3 Aug 16 2008 SM - changed baud rate to 10.417K */ Copyright Freescale, 2007. All Rights Reserved */ Notes: 1 MMU not initialized; must be done by debug scripts or BAM */ 2 SRAM not initialized; must be done by debug scripts or in a startup file*/ #include "mpc563m.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ const uint8_t FrameSpecAndData[]={0x35,0x08,0xD0,'H','e','l','l','o',' ',' ',' '}; void initSysclk (void) { /* MPC551x: Use the next 6 lines */ /* CRP.CLKSRC.B.XOSCEN = 1; */ /* Enable external oscillator */ /* FMPLL.ESYNCR2.R = 0x00000006; */ /* Set ERFD to initial value of 6 */ /* FMPLL.ESYNCR1.R = 0xF0000020; */ /* Set CLKCFG=PLL, EPREDIV=0, EMFD=0x20*/ /* while (FMPLL.SYNSR.B.LOCK != 1) {};*/ /* Wait for PLL to LOCK */ /* FMPLL.ESYNCR2.R = 0x00000005; */ /* Set ERFD to final value for 64 MHz sysclk /* SIU.SYSCLK.B.SYSCLKSEL = 2; */ /* Select PLL for sysclk */ /* MPC563x: Use the next line */ FMPLL.ESYNCR1.B.CLKCFG = 0X7; /* Change clk to PLL normal mode from crystal /* MPC555x including MPC563x: use the next 3 lines for either 8 or 40 MHz crystal FMPLL.SYNCR.R = 0x16080000; /* 8 MHz xtal: 0x16080000; 40MHz: 0x46100000 while (FMPLL.SYNSR.B.LOCK != 1) {}; /* Wait for FMPLL to LOCK */ FMPLL.SYNCR.R = 0x16000000; /* 8 MHz xtal: 0x16000000; 40MHz: 0x46080000 } */ */ */ */ */ void initESCI_B (void) { ESCI_B.CR2.R = 0x6240; /* Module is enabled, 13 bit break, stop on errors */ ESCI_B.CR1.R = 0x0180000C; /* 10417 baud, 8 bits, no parity, Tx & Rx enabled */ ESCI_B.LCR.R = 0x01000000; /* eSCI put in LIN mode */ /* Use the following two lines for MPC551x */ /* SIU.PCR[56].R = 0x400; */ /* Configure pad for primary func: TxDB */ /* SIU.PCR[57].R = 0x400; */ /* Configure pad for primary func: RxDB */ /* Use the following two lines for MPC555x */ SIU.PCR[91].R = 0x400; /* Configure pad for primary func: TxDB */ SIU.PCR[92].R = 0x400; /* Configure pad for primary func: RxDB */ } void TransmitData (void) { uint8_t j; /* Dummy variable */ for (j=0; j< sizeof (FrameSpecAndData); j++) { /* Loop for character string */ while (ESCI_A.SR.B.TXRDY == 0) {} /*Wait for LIN transmit ready = 1*/ ESCI_A.SR.R = 0x00004000; /* Clear TXRDY flag */ ESCI_A.LTR.R = FrameSpecAndData[j]; /* Write 8 byte to LIN Trans Reg.*/ } } void main(void) { initSysclk(); initESCI_B(); TransmitData(); while (1) {} } /* /* /* /* Set sysclk = 64MHz running from PLL */ Enable Tx for 10417 baud, 8 bits, no parity */ Transmit string of characters on eSCI B */ Wait forever */ Qorivva Simple Cookbook, Rev. 4 138 Freescale Semiconductor 17 LINFlex: LIN Transmit 17.1 Description Task: Transmit the data string “Hello” and three spaces using LIN. Use parameters as in the table below. Table 63. MPC56xxB/P/S LINFlex — LIN example parameters Parameter General Frame Register[Bit field] Setting Master or Slave Master LINCR1[MME] Master break length 13-bit break LINCR1[MBL] CKSUM done in HW Yes (no CRC is used) LINCR1[CCD] CKSUM enabled Yes LINCR1[CFD] Baud rate 10417 Hz for 64 MHz clock LINIBRR, LINFBRR Data String “Hello” and 3 spaces BDRM, BDRL ID 0x35 Data field length 8 bytes BIDR[DFL] Message direction Transmit BIDR[DIR] CKSYM type Enhanced (LIN 2.0) BIDR[CCS] BIDR[ID] NOTE The RxD receive pin must be able to monitor the TxD transmitted signal. Hence either (1) the LIN transceiver must be connected as shown and the transceiver powered, or (2) for debug purpuses, the TxD and RxD can be shorted together without using the transceiver. Exercise: Connect an oscilloscope or LIN tool and verify the transmit output. MPC56xxB/P/S Clock Generation Module Crystal OSC0 64 MHz sysclk LINFlex TxD LIN Control LIN Status Baud rate Filter config. Message Buffer Interface RxD XCVR LIN Figure 30. LINFlex LIN Transmit Example Simplified Block Diagram Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 139 Table 64. MPC56xxB/P/S Signals for LINFlex LIN TransmitScan Example Signal MPC56xxB Family Port MPC56xxP Family SIU Pad Package Pin # Configuration & Selection 100 144 176 Registers QFP QFP BGA (values in hex) TxD B2 PCR18=0400 LIN0 TX RxD B3 PCR19=0103 LIN0 no PSMI RX 100 144 1 1 Port MPC56xxS Family SIU Pad Package Port Configuration Pin # & Selection 100 144 Registers QFP QFP (values in hex) SIU Pad Package Pin # Configuration & Selection 144 176 208 Registers QFP QFP BGA (values in hex) B2 B2 PCR18=0400 TXD 79 114 B2 PCR18=0400 TxD_ A 112 140 R14 B3 B3 PCR19=0503 RXD PSMI31 = 0 80 116 B3 PCR19=0503 RxD_ A 111 139 R13 Qorivva Simple Cookbook, Rev. 4 140 Freescale Semiconductor 17.2 Design 17.2.1 Mode Use Mode Transition is required for changing mode entry registers. Hence even enabling the crystal oscillator to be active in the current mode (for example default mode (DRUN)) requires enabling the crystal oscillator in DRUN mode configuration register (ME_DRUN_MC), then initiating a mode transition to the same DRUN mode. This example changes from DRUN mode to RUN0 mode. This minimal example simply polls a status bit to wait for the targeted mode transition to complete. However, the status bit could instead be enabled to generate an interrupt request (assuming the INTC is intialized beforehand). This would allow software to complete other intialization tasks instead of brute force polling of the status bit. It is normal to use a timer when waiting for a status bit to change. This example by default would have a watchdog timer expire if for some reason the mode transition never completed. One could also loop code on incrementing a software counter to some maximum value as a timeout. If a timeout was reached, then an error condition could be recorded in EEPROM or elsewhere. Table 65. Mode Configurations Summary for MPC56xxB/P/S LINFlex LIN Transmit Example (modes are enabled in ME_ME register) Settings Mode Mode Config. Register Value sysclk Selection DRUN ME_DRUN_MC 0x001F 0010 (default) RUN0 Memory Power Mode Clock Sources Mode Config. Register ME_RUN0_MC 0x001F 0074 PLL1 16 MHz XOSC0 PLL0 (MPC Data IRC 56xxP/S Flash only) Code Flash Main Voltage Reg. I/O Power Down Ctrl 16 MHz IRC On Off Off Off Normal Normal On Off PLL0 On On On Off Nomral Normal On Off Other modes are not used in example. Peripherals also have configurations to gate clocks on and off, enabling low power. The following table summarizes the peripheral configurations used here. ME_RUNPC_1 is selected, therefore peripherals to be used require a non-zero value in their respective ME_PCTL register. Table 66. Peripheral Configurations for MPC56xxB/P/S LINFlex LIN Transmit Example (low power modes are not used in example) PeriPeri. Enabled Modes pheral Config. Config. Register RUN3 RUN2 RUN1 RUN0 DRUN SAFE TEST RESET PC1 ME_ RUNPC_ 1 0 0 0 1 0 0 0 0 Peripherals Selecting Configuration Peripheral PCTL Reg. # LINFlex 0 LINFlex 1 SIUL (MPC56xxB/S only) 48 29 68 Other peripheral configurations are not used in example Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 141 17.2.2 Steps and Pseudo Code Table 67. MPC5606B, MPC56xxP, MPC56xxS Steps for LINFlex LIN Transmit Example Pseudo Code Step Relevant Bit Fields MPC56xxB Init Modes and Clock Enable desired modes RUN0, DRUN=1 Initialize PLL0 dividers to provide 64 MHz for input crystal before mode configuration: • 8 MHz xtal: FMPLL[0]_CR=0x02400100 • 40 MHz xtal: FMPLL[0]_CR=0x12400100 (MPC56xxP & 8 MHz xtal requires changing default CMU_CSR value. See PLL example.) MPC56xxS ME_ME = 0x0000 001D CGM_FMPLL_CR (MPC56xxB) CGM_FMPLL[0]_CR (MPC56xxS) = 0x0240 0100 (8 MHz crystal) or 0x1240 0100 (40 MHz crystal) Configure RUN0 Mode: • I/O Output power-down: no safe gating • Main Voltage regulator is on (default) • Data, code flash in normal mode (default) • PLL0 is switched on • Crystal oscilator is switched on • 16 MHz IRC is switched on (default) • Select PLL0 (system pll) as sysclk PDO=0 MVRON=1 DFLAON = 3 CFLAON = 3 PLL0ON=1 XOSC0ON=1 16MHz_IRCON = 1 SYSCLK=0x4 ME_RUN0_MC = 0x001F 0074 Peri. Config. 1: run in RUN0 mode only RUN0=1 ME_RUN_PC1 = 0x0000 0010 Assign peripheral configuration to peripherals: • LINFlex 0: select ME_RUN_PC0 • LINFlex 1: select ME_RUN_PC0 • SIUL: select ME_RUN_PC0 (56xxB/S) RUN_CFG = 1 RUN_CFG = 1 RUN_CFG = 1 Initiate software mode transition to RUN0 mode • Mode and key • Mode and inverted key • Wait for mode transition complete status flag • Clear transition complete status flag NOTE: if transition does not complete, check status flags such as ME_GS[XOSC] • Verify desired target mode was entered Disable • Write keys to clear soft lock bit Watchdog • Clear watchdog enable bit init Peri Clk Gen MPC56xxP TARGET_MODE= RUN0 wait for I_TC = 1 I_TC=1 WEN = 0 Initialize peripheral clockgeneration (See appendix: MPC56xxB/P/S Peripheral Clocks) DE0=1 • LINFlex (56xxB/S): peripheral set 1 – sysclk/1 DIV0=0 ME_PCTL48 = 0x01 ME_PCTL49 = 0x01 .ME_PCTL68 = 0x01 (56xxB/S only) ME_MCTL =0x4000 5AF0 ME_MCTL =0x4000 A50F wait for ME_IS[I_TC] = 1 ME_IS[I_MTC] = 1 verify ME_GS[S_CURRENT_MODE] = RUN0 SWT_SR = 0x000 0C520 SWT_SR = 0x0000 D928 SWT_CR = 0x8000 010A CGM_ SC_DC0 = 0x80 - CGM_ SC_DC0 = 0x80 Qorivva Simple Cookbook, Rev. 4 142 Freescale Semiconductor Table 67. MPC5606B, MPC56xxP, MPC56xxS Steps for LINFlex LIN Transmit Example Pseudo Code Step Relevant Bit Fields MPC56xxB Init Put LINFlex hardware in init mode LINFlex_0 Initialize module: (as LIN • Mode = LIN Master Master) • LIN master break length = 13 • CKSUM done in hardware • CKSUM field (for LIN frame) enabled • Module HW put in INIT mode INIT = 1 LINIBRR = 384 LINFBRR = 0 INIT = 0 SLEEP = 0 LINCR1 = 0x0000 0310 SIU_PCR18 = SIU_PCR18 = SIU_PCR18 = 0x0400 0x0400 0x0400 SIU_PCR19 = SIU_PCR19= SIU_PCR19= 0x0103 0x0503 0x0503 PSMI31=0 Load buffer data with 8 bytes: “Hello” and 3 blank spaces (hex 48 65 6C 6C 6F 20 20 20) Init header in Buffer ID register: • ID = 0x35 • Data field length = 8 bytes • Message direction = transmit • Cksum tyipe = Enhanced (for LIN 2.0) BDRM = 0x4865 6C6C BDRL = 0x6F20 2020 BIDR = 0x0000 1E35 ID = 0x35 DFL = 7 DIR = 1 CCS = 0 Request header transmission (using default error HTRQ=1 controls) 1 LINCR1[INIT] = 1 LINCR1 = 0x0000 0311 Init pads for FlexLIN_0 TxD & RxD Transmit LIN Frame MPC56xxS MME = 1 MBL = 3 CCD = 0 CFD = 0 INIT = 1 Set baud rate = 10417 for 64 MHz sysclk1 Put LINFlex hardware in normal mode MPC56xxP LINCR2[HTRQ] = 1 Per MPC5606S Microcontroller Reference Manual Rev 3 Table 22-1 for 10417 baud rate. Note: LINFBRR[DIV_F] field is 4 bits, so the value must be less than 16. Some documentation has a table incorrectly showing value of 16 for this field at 10417 baud. Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 143 17.3 17.3.1 Code MPC560xB /* main.c: LINFlex program for MPC56xxB */ /* Description: Transmit one message from FlexCAN 0 buf. 0 to FlexCAN C buf. 1 */ /* Oct 30 2009 SM - initial version */ /* Mar 14 2010 SM - modified initModesAndClock, updated header file */ #include "MPC5604B_0M27V_0101.h" /* Use proper header file*/ void initModesAndClks(void) { ME.MER.R = 0x0000001D; /* Enable DRUN, RUN0, SAFE, RESET modes */ /* Initialize PLL before turning it on: */ /* Use 1 of the next 2 lines depending on crystal frequency: */ CGM.FMPLL_CR.R = 0x02400100; /* 8 MHz xtal: Set PLL0 to 64 MHz */ /*CGM.FMPLL_CR.R = 0x12400100;*/ /* 40 MHz xtal: Set PLL0 to 64 MHz */ ME.RUN[0].R = 0x001F0074; /* RUN0 cfg: 16MHzIRCON,OSC0ON,PLL0ON,syclk=PLL */ ME.RUNPC[1].R = 0x00000010; /* Peri. Cfg. 1 settings: only run in RUN0 mode */ ME.PCTL[48].R = 0x01; /* MPC56xxB/P/S LINFlex 0: select ME.RUNPC[1] */ ME.PCTL[68].R = 0x01; /* MPC56xxB/S SIUL: select ME.RUNPC[0] */ /* Mode Transition to enter RUN0 mode: */ ME.MCTL.R = 0x40005AF0; /* Enter RUN0 Mode & Key */ ME.MCTL.R = 0x4000A50F; /* Enter RUN0 Mode & Inverted Key */ while (ME.GS.B.S_MTRANS) {} /* Wait for mode transition to complete */ /* Note: could wait here using timer and/or I_TC IRQ */ while(ME.GS.B.S_CURRENTMODE != 4) {} /* Verify RUN0 is the current mode */ } void initPeriClkGen(void) { CGM.SC_DC[0].R = 0x80; /* MPC56xxB/S: Enable peri set 1 sysclk divided by 1 */ } void disableWatchdog(void) { SWT.SR.R = 0x0000c520; /* Write keys to clear soft lock bit */ SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; /* Clear watchdog enable (WEN) */ } void initLINFlex_0 (void) { LINFLEX_0.LINCR1.B.INIT = 1; LINFLEX_0.LINCR1.R= 0x00000311; LINFLEX_0.LINIBRR.B.DIV_M= 383; LINFLEX_0.LINFBRR.B.DIV_F = 16; LINFLEX_0.LINCR1.R= 0x00000310; SIU.PCR[18].R = 0x0400; SIU.PCR[19].R = 0x0103; } void transmitLINframe (void) { LINFLEX_0.BDRM.R = 0x2020206F; LINFLEX_0.BDRL.R = 0x6C6C6548; LINFLEX_0.BIDR.R = 0x00001E35; LINFLEX_0.LINCR2.B.HTRQ = 1; } void main(void) { volatile uint32_t IdleCtr = 0; } initModesAndClks(); initPeriClkGen(); disableWatchdog(); initLINFlex_0(); transmitLINframe(); while (1) {IdleCtr++;} /* /* /* /* /* /* /* Put LINFlex hardware in init mode */ Configure module as LIN master & header */ Mantissa baud rate divider component */ Fraction baud rate divider comonent */ Configure module as LIN master & header */ MPC56xxB: Configure port B2 as LIN0TX */ MPC56xxB: Configure port B3 as LIN0RX */ /* /* /* /* Load buffer data most significant bytes */ Load buffer data least significant bytes */ Init header: ID=0x35, 8 B, Tx, enh. cksum*/ Request header transmission */ /* Initialize mode entries */ /* Initialize peripheral clock generation for LINFlex */ /* Disable watchdog */ /* Initialize FLEXCAN 0 as master */ /* Transmit one frame from master */ /* Idle loop: increment counter */ Qorivva Simple Cookbook, Rev. 4 144 Freescale Semiconductor 17.3.2 /* /* /* /* MPC560xP main.c: LINFlex program for MPC56xxP */ Description: Transmit one message from FlexCAN 0 buf. 0 to FlexCAN C buf. 1 */ Rev Oct 30 2009 SM - initial version */ Rev Mar 14 1020 SM - Modified initModesAndClks, updated header */ #include "jdp_pictus_0106.h" /* Use proper include file */ void initModesAndClks(void) { ME.MER.R = 0x0000001D; /* Enable DRUN, RUN0, SAFE, RESET modes */ /* Initialize PLL before turning it on: */ /* Use 2 of the next 4 lines depending on crystal frequency: */ /*CGM.CMU_0_CSR.R = 0x000000004;*//* Monitor FXOSC > FIRC/4 (4MHz); no PLL monitor */ /*CGM.FMPLL[0].CR.R = 0x02400100;*/ /* 8 MHz xtal: Set PLL0 to 64 MHz */ CGM.CMU_0_CSR.R = 0x000000000; /* Monitor FXOSC > FIRC/1 (16MHz); no PLL monitor*/ CGM.FMPLL[0].CR.R = 0x12400100; /* 40 MHz xtal: Set PLL0 to 64 MHz */ ME.RUN[0].R = 0x001F0074; /* RUN0 cfg: 16MHzIRCON,OSC0ON,PLL0ON,syclk=PLL */ ME.RUNPC[1].R = 0x00000010; /* Peri. Cfg. 1 settings: only run in RUN0 mode */ ME.PCTL[48].R = 0x01; /* MPC56xxB/P/S LINFlex 0: select ME.RUNPC[1] */ /* Mode Transition to enter RUN0 mode: */ ME.MCTL.R = 0x40005AF0; /* Enter RUN0 Mode & Key */ ME.MCTL.R = 0x4000A50F; /* Enter RUN0 Mode & Inverted Key */ while (ME.GS.B.S_MTRANS) {} /* Wait for mode transition to complete */ /* Note: could wait here using timer and/or I_TC IRQ */ while(ME.GS.B.S_CURRENTMODE != 4) {} /* Verify RUN0 is the current mode */ } void initPeriClkGen(void) { } void disableWatchdog(void) { SWT.SR.R = 0x0000c520; /* Write keys to clear soft lock bit */ SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; /* Clear watchdog enable (WEN) */ } void initLINFlex_0 (void) { LINFLEX_0.LINCR1.B.INIT = 1; LINFLEX_0.LINCR1.R= 0x00000311; LINFLEX_0.LINIBRR.B.DIV_M= 383; LINFLEX_0.LINFBRR.B.DIV_F = 16; LINFLEX_0.LINCR1.R= 0x00000310; SIU.PCR[18].R = 0x0400; SIU.PCR[19].R = 0x0503; SIU.PSMI[31].R = 0; } void transmitLINframe (void) { LINFLEX_0.BDRM.R = 0x2020206F; LINFLEX_0.BDRL.R = 0x6C6C6548; LINFLEX_0.BIDR.R = 0x00001E35; LINFLEX_0.LINCR2.B.HTRQ = 1; } /* /* /* /* /* /* /* /* Put LINFlex hardware in init mode */ Configure module as LIN master & header */ Mantissa baud rate divider component */ Fraction baud rate divider comonent */ Configure module as LIN master & header */ MPC56xxP: Configure port B2 as LIN0TX */ MPC56xxP: Configure port B3 as LIN0RX */ MPC56xxP: LIN0 Pad select mux port B3 */ /* /* /* /* Load buffer data most significant bytes */ Load buffer data least significant bytes */ Init header: ID=0x35, 8 B, Tx, enh. cksum*/ Request header transmission */ void main(void) { volatile uint32_t IdleCtr = 0; } initModesAndClks(); initPeriClkGen(); disableWatchdog(); initLINFlex_0(); transmitLINframe(); while (1) {IdleCtr++;} /* Initialize mode entries */ /* Initialize peripheral clock generation for LINFlex */ /* Disable watchdog */ /* Initialize FLEXCAN 0 as master */ /* Transmit one frame from master */ /* Idle loop: increment counter */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 145 17.3.3 /* /* /* /* MPC56xxS main.c: LINFlex program for MPC56xxS*/ Description: Transmit one message from FlexCAN 0 buf. 0 to FlexCAN C buf. 1 */ Oct 30 2009 SM - initial version */ Mar 15 2010 SM - modified initModesAndClks, updated header */ #include "56xxS_0204.h" /* Use proper header file */ void initModesAndClks(void) { ME.MER.R = 0x0000001D; } /* Enable DRUN, RUN0, SAFE, RESET modes */ /* Initialize PLL before turning it on: */ CGM.FMPLL[0].CR.R = 0x02400100; /* 8 MHz xtal: Set PLL0 to 64 MHz */ ME.RUN[0].R = 0x001F0074; /* RUN0 cfg: 16MHzIRCON,OSC0ON,PLL0ON,syclk=PLL */ ME.RUNPC[1].R = 0x00000010; /* Peri. Cfg. 1 settings: only run in RUN0 mode */ ME.PCTL[48].R = 0x01; /* MPC56xxB/P/S LINFlex 0: select ME.RUNPC[1] */ ME.PCTL[68].R = 0x01; /* MPC56xxB/S SIUL: select ME.RUNPC[1] */ /* Mode Transition to enter RUN0 mode: */ ME.MCTL.R = 0x40005AF0; /* Enter RUN0 Mode & Key */ ME.MCTL.R = 0x4000A50F; /* Enter RUN0 Mode & Inverted Key */ while (ME.GS.B.S_MTRANS) {} /* Wait for mode transition to complete */ /* Note: could wait here using timer and/or I_TC IRQ */ while(ME.GS.B.S_CURRENTMODE != 4) {} /* Verify RUN0 is the current mode */ void initPeriClkGen(void) { CGM.SC_DC[0].R = 0x80; } /* MPC56xxB/S: Enable peri set 1 sysclk divided by 1 */ void disableWatchdog(void) { SWT.SR.R = 0x0000c520; /* Write keys to clear soft lock bit */ SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; /* Clear watchdog enable (WEN) */ } void initLINFlex_0 (void) { } LINFLEX_0.LINCR1.B.INIT = 1; LINFLEX_0.LINCR1.R= 0x00000311; LINFLEX_0.LINIBRR.B.DIV_M= 383; LINFLEX_0.LINFBRR.B.DIV_F = 16; LINFLEX_0.LINCR1.R= 0x00000310; SIU.PCR[18].R = 0x0400; SIU.PCR[19].R = 0x0503; void transmitLINframe (void) { LINFLEX_0.BDRM.R = 0x2020206F; LINFLEX_0.BDRL.R = 0x6C6C6548; LINFLEX_0.BIDR.R = 0x00001E35; LINFLEX_0.LINCR2.B.HTRQ = 1; } /* /* /* /* /* /* /* Put LINFlex hardware in init mode */ Configure module as LIN master & header */ Mantissa baud rate divider component */ Fraction baud rate divider comonent */ Configure module as LIN master & header */ MPC56xxS: Configure port B2 as LIN0TX */ MPC56xxS: Configure port B3 as LIN0RX */ /* /* /* /* Load buffer data most significant bytes */ Load buffer data least significant bytes */ Init header: ID=0x35, 8 B, Tx, enh. cksum*/ Request header transmission */ void main(void) { volatile uint32_t IdleCtr = 0; } initModesAndClks(); initPeriClkGen(); disableWatchdog(); initLINFlex_0(); transmitLINframe(); while (1) {IdleCtr++;} /* /* /* /* /* /* Initialize mode entries */ Initialize peripheral clock generation for LINFlex */ Disable watchdog */ Initialize FLEXCAN 0 as master */ Transmit one frame from master */ Idle loop: increment counter */ Qorivva Simple Cookbook, Rev. 4 146 Freescale Semiconductor 18 eMIOS: Modulus Counter, OPWM Functions 18.1 Description Task: Provide two Output Pulse Width Modulation (OPWM) signals that are synchronized to a common counter bus. The common counter bus is a separate eMIOS channel, configured as a modulus counter. The original modulus counter and OPWM modes in early MPC555x devices were later replaced by modulus counter buffered and OPWM buffered modes in some MPC555x devices and in all MPC551x devices. Code is provided to manage both cases. The MPC563x lacks OPWMB mode on channel 1, so channel 2 is used in the example code. MPC56xxB/S use eMIOS_0. Exercise: If using the MPC5500 Evaluation Board, connect an eMIOS output to the speaker (if available) and/or an LED or oscilloscope. Observe the outputs, then alter code to change the frequency or duty cycle. Alternate exercise: Wire an eMIOS channel to an LED, and slow down the oscillations using code changes below. Observe pulses on LED. 1. Increase eMIOS Prescaler. Example: EMIOS_0.MCR.B.GPRE= 254; 2. Increase Channel 23 Modulus Counter value. Example: EMIOS_0.CH[23].CADR.R = 64000; 3. Increase Channel 21 falling edge value. Example: EMIOS_0.CH[21].CBDR.R = 32000; MPC5500 / MPC5600 8 MHz Crystal Clocks and PLL sysclk = 64 MHz eMIOS (eMIOS_0 on MPC56xxB/P/S) eMIOS Prescaler (Divide by 64) eMIOS Channel 23 (Mod. Counter: divide by 1 & count 1000 clocks) 1 MHz eMIOS internal clock (available to all eMIOS channels) eMIOS Channel 0 or 21 (OPWM or OPWMB: Rising edge at 250 Falling edge at 500) eMIOS Channel 1 or 2 or 22 (OPWM or OPWMB: Rising edge at 500 Falling edge at 999) EMIOS[0] or EMIOS_0[21] EMIOS[1] or EMIOS[2] or EMIOS_0[22] Time Base A (Counts 1000 1 µsec clocks; available to all eMIOS channels) Figure 31. eMIOS OPWM Example Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 147 Table 68. MPC551x, MPC55xx Signals for eMIOS OPWM Example MPC551x Family Signal Pin Name SIU PCR No. MPC555x Family Package Pin No. 144 QFP 176 QFP 208 BGA Function Name SIU PCR No. Package Pin No. EVB 496 BGA 416 BGA 324 BGA 208 BGA xPC 563M EMIOS[0] PC0 32 122 146 B11 EMIOS[0] 179 AD17 AF15 AB10 T4 PJ8–1 EMIOS[1] PC1 33 121 145 C11 EMIOS[1] 180 AD21 AE15 AB11 T5 n.a. EMIOS[2] 181 EMIOS[2] not used in MPC551x code only used in MPC563x code PJ8–2 Table 69. MPC56xxB/S Signals for eMIOS OPWM Example Signal MPC56xxB Family Port MPC56xxS Family SIU Pad Package Pin # Port SIU Pad Package Pin # Configuration Configuration 100 144 176 144 176 208 & Selection & Selection QFP QFP BGA QFP QFP BGA Registers Registers (values in hex) (values in hex) EMIOS_0[21] E5 PCR69=0600 94 133 C6 A[1] PCR1=0A00 EMIOS_0[22] E6 PCR70=0600 95 139 B5 A[0] PCR0=0A00 136 166 B1 135 165 A1 Qorivva Simple Cookbook, Rev. 4 148 Freescale Semiconductor 18.2 Design Timing resources used will include: • sysclk: 64 MHz — assume 8 MHz crystal unless noted • eMIOS internal clock: Choose 1 MHz (requires prescaling sysclk by 64) • eMIOS Channel 23: initialize in modulus counter mode, which will be used as the global counter bus, up-counting 1000 eMIOS internal clocks (use value of 1000 – 1 = 999) • eMIOS Channels 0 and 1: OPWM mode based on Time Bus A, each channel with different duty cycles; the signal polarity will be rising edge for the first match, falling edge for the second match 18.2.1 Mode Use Summary (MPC56xxB/S only) Mode Transition is required for changing mode entry registers. Hence even enabling the crystal oscillator to be active in the default mode (DRUN) requires enabling the crystal oscillator in appropriate mode configuration register (ME_xxxx_MC) then initiating a mode transition. This example transitions from the default mode after reset (DRUN) to RUN0 mode. Table 70. Mode Configurations for MPC56xxB/S DSPI SPI to SPI Example Modes are enabled in ME_ME Register. Settings Memory Power Mode Clock Sources Mode Mode Config. Register Mode Config. Register Value sysclk Selection PLL1 16MHz XOSC0 PLL0 (MPC IRC 56xxP/S only) Data Flash Code Flash Main Voltage Reg. I/O Power Down Ctrl DRUN ME_DRUN_MC 0x001F 0010 16 MHz IRC (default) On Off Off Off Normal Normal On Off RUN0 On On On Off Normal Normal On Off ME_RUN0_MC 0x001F 007D PLL0 Other modes are not used in example Peripherals also have configurations to gate clocks on or off for different modes, enabling low power. The following table summarizes the peripheral configurations used in this example. Table 71. Peripheral Configurations for MPC56xxB/S DSPI SPI to SPI Example Low power modes are not used in example. PeriPeri. Enabled Modes pheral Config. Config. Register RUN3 RUN2 RUN1 RUN0 DRUN SAFE TEST RESET PC1 ME_ RUNPC_ 1 0 0 0 1 0 0 0 0 Peripherals Selecting Configuration Peripheral SIUL (MPC56xxB/S) eMIOS 0 PCTL Reg. # 68 72 Other peripheral configurations are not used in example Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 149 18.2.2 Design Steps Table 72. eMIOS Modulus Counter and Output Pulse Width (OPWM) Example Pseudo Code Step Relevant Bit Field MPC551x init Enable desired modes Modes and Clock Initialize PLL0 dividers to provide 64 MHz for input crystal before mode configuration: (MPC • 8 MHz xtal: FMPLL[0]_CR=0x02400100 56xxPBS • 40 MHz xtal: FMPLL[0]_CR=0x12400100 only) (MPC56xxP & 8 MHz xtal requires changing default CMU_CSR value. See PLL example.) DRUN=1, RUN0 = 1 MPC56xxB/S - ME_ME = 0x0000 001D - 8 MHz Crystal: CGM_ FMPLL[0]_CR =0x02400100 - ME_ RUN0_MC = 0x001F 0070 - ME_RUN_PC1 = 0000 0010 Assign peripheral configuration to peripherals: • SIUL: select ME_RUN_PC1 (MPC56xxB/S) RUN_CFG = 1 • eMIOS 0: select ME_RUN_PC1 RUN_CFG = 1 - MC_PCTL68 = MC_PCTL72 = 0x01 Initiate software mode transition to RUN0 mode • Mode & key, then mode & inverted key TARGET_MODE = • Wait for transition to complete RUN0 S_TRANS • Verify current mode is RUN0 CURRENTMODE - ME_MCTL =0x4000 5AF0, =0x4000 A50F wait ME_GS [S_TRANS] = 0 verify 4 = ME_GS [CURRENTMODE] See PLL Initialization example - - MPC56xxB: CGM_SC_DC= 0x0000 8000 MPC56xxS: CGM_AC_DC2= 0x80 - See PLL Initialization example Configure RUN0 Mode: • I/O Output power-down: no safe gating • Main Voltage regulator is on (default) • Data, code flash in normal mode (default) • PLL0 is switched on • Crystal oscillator is switched on • 16 MHz IRC is switched on (default) • Select PLL0 (system pll) as sysclk PDO=0 MVRON=1 DFLAON, CFLAON= 3 PLL0ON=1 XOSC0ON=1 16MHz_IRCON=1 SYSCLK=0x4 MPC56xxB/S: • Peri. Config. 1: run in RUN0 mode only RUN0=1 init Sysclk Initialize sysclk to 64 MHz, running from PLL init Peri Clk Gen (MPC 56xxPBS only) MPC555x Initialize peripheral clock generation (See appendix: MPC56xxB/P/S Peripheral Clocks) • MPC56xxB: Enable Peri. Set 2- sysclk div. by 1 • MPC56xxS: Enable Aux Clk 1- PLL div. by 1 disable Disable watchdog by writing keys to Status Watchdog Register, then clearing WEN (MPC56xxBPS (56xxPBS) only) Qorivva Simple Cookbook, Rev. 4 150 Freescale Semiconductor Table 72. eMIOS Modulus Counter and Output Pulse Width (OPWM) Example (continued) Pseudo Code Step Relevant Bit Field MPC551x initEMIOS Config eMIOS for 1 MHz internal clock: • Set global prescaler = divide by 64 (63 + 1) • Do not use external time base • Enable global prescaler & internal clock • Enable global time base • Enable freezing counters during debug init EMIOS Set Channel’s A match data for 1000 counts ch 23 (Match value used for modulus counter) GPRE = 63(0x3F) ETB = 0 (default) GPREN = 1 GTBE = 1 FRZ = 1 A = 999 MPC555x MPC56xxB/S EMIOS_ MCR = 0x3400 3F00 EMIOS_0_ MCR = 0x3400 3F00 EMIOS_ CH[23]CADR = 999 EMIOS_0_ CH[23]CADR = 999 EMIOS_ EMIOS_ MODE= 0x50(551x) CCR[23] = CCR[23] = MODE= 0x10(555x) 0x8202 0650 0x8202 0610 BSL = 3 UCPRE= 0 UCPREN = 1 FREN = 1 EMIOS_0_ CCR[23] = 0x8202 0650 init EMIOS Set Channel’s A match Data value = 250 ch 0 A = 250 EMIOS_ CH[0]CADR = 250 EMIOS_0_ CH[0]CADR = 250 Set Channel’s B match Data value = 500 B = 500 EMIOS_ CH[0]CBDR = 500 EMIOS_0_ CH[0]CBDR = 500 Set up Channel Control: • BSL = Bus selected is counter bus A • EDPOL = leading edge sets; trailing clears • Mode (551x, 563x, 56xxB/S) = Output PWMB • Mode (MPC555x) = Output PWM EMIOS_ BSL=0 (default) CH[0]CCR = EDPOL=1 0x0000 Mode=0x60 (551x 00E0 or 563x) Mode=0x20 (555x) Configure pad for eMIOSchannel 0: • Pad assignment= EMIOS Ch 0 • Pad output buffer enabled PA = 1 (MPC551x) SIU_PCR[32] SIU_PCR[179] = 0x0600 = 0x0E00 PA = 3 (MPC555x) OBE = 1 Channel 23: Enable channel as up counter • Mode (551x, 56xxB/S) = mod. up counter buf’d • Mode (555x) = mod. up counter • Counter bus select = internal counter • Channel prescaler = 1 • Enable channel prescaler • Enable freezing count in debug mode init EMIOS Set Channel’s A match Data value = 500 ch 1 (MPC563x :channel 2 is used because channel 1 lacks OPWMB mode) EMIOS_ CH[0]CCR = 0x0000 00A0 or (MPC563x): 0x0000 00E0 EMIOS_0_ CH[0]CCR = 0x0000 00E0 See table: MPC56xxB/P/S Signals A = 500 EMIOS_ CH[1]CADR = 500 EMIOS_0_ CH[1]CADR = 500 Set Channel’s B match Data value = 999 B = 999 EMIOS_ CH[1]CBDR = 999 EMIOS_0_ CH[1]CBDR = 999 Set up Channel Control: • BSL = Bus selected is counter bus A • EDPOL = leading edge sets; trailing clears • Mode (551x, 563x, 56xxB/S) = Output PWMB • Mode (MPC555x) = Output PWM EMIOS_ EMIOS_ BSL=0 (default) CH[1]CCR = CH[1]CCR = EDPOL=1 0x0000 0x0000 00A0 Mode=0x60 (551x 00E0 or (MPC563x): or 563x) 0x0000 00E0 Mode=0x20 (555x) EMIOS_0_ CH[1]CCR = 0x0000 00E0 Configure pad for eMIOS channel 1: • Pad assignment = EMIOS Ch 1 • Pad output buffer enabled PA = 1 (MPC551x) SIU_PCR[33] SIU_PCR[180] = 0x0600 = 0x0E00 PA = 3 (MPC555x) OBE = 1 See table: MPC56xxB/S Signals Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 151 18.2.3 Design Screenshot The screenshot below shows the two OPWM channels as a result of this design. Both channels schedule leading and trailing edges based on the counter of channel 23, which counts to 1000 s. Hence the OPWM frequency is 1 kHz. Qorivva Simple Cookbook, Rev. 4 152 Freescale Semiconductor 18.3 18.3.1 /* /* /* /* /* /* /* /* /* /* /* /* Code MPC551x, MPC555x main.c - eMIOS OPWM example */ Description: eMIOS example using Modulus Counter and OPWM modes */ Rev 1.0 Sept 9 2004 S.Mihalik */ Rev 1.1 April 13 2006 S.M.- corrected GPRE to be div by 12 instead of 13*/ Rev 1.2 June 26 1006 S.M. - updated comments & made i volatile uint32_t */ Rev 1.3 July 19 2007 SM- Changes for MPC551x, 50 MHz sysclk, Mod Ctr data value*/ Rev 1.4 Aug 10 2007 SM - Changed to use sysclk of 64 MHz */ Rev 1.5 Jun 04 2008 SM - initSysclk changed for MPC5633M support */ Copyright Freescale Semiconductor, Inc. 2007 All rights reserved. */ Notes: */ 1. MMU not initialized; must be done by debug scripts or BAM */ 2. SRAM not initialized; must be done by debug scripts or in a crt0 type file */ #include "mpc5554.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ void initSysclk (void) { /* MPC551x: Use the next 6 lines */ /* CRP.CLKSRC.B.XOSCEN = 1; */ /* Enable external oscillator */ /* FMPLL.ESYNCR2.R = 0x00000006; */ /* Set ERFD to initial value of 6 */ /* FMPLL.ESYNCR1.R = 0xF0000020; */ /* Set CLKCFG=PLL, EPREDIV=0, EMFD=0x20*/ /* while (FMPLL.SYNSR.B.LOCK != 1) {};*/ /* Wait for PLL to LOCK */ /* FMPLL.ESYNCR2.R = 0x00000005; */ /* Set ERFD to final value for 64 MHz sysclk /* SIU.SYSCLK.B.SYSCLKSEL = 2; */ /* Select PLL for sysclk */ /* MPC563x: Use the next line */ /* FMPLL.ESYNCR1.B.CLKCFG = 0X7; */ /* Change clk to PLL normal mode from crystal /* MPC555x including MPC563x: use the next 3 lines for either 8 or 40 MHz crystal FMPLL.SYNCR.R = 0x16080000; /* 8 MHz xtal: 0x16080000; 40MHz: 0x46100000 while (FMPLL.SYNSR.B.LOCK != 1) {}; /* Wait for FMPLL to LOCK */ FMPLL.SYNCR.R = 0x16000000; /* 8 MHz xtal: 0x16000000; 40MHz: 0x46080000 } */ */ */ */ */ void initEMIOS(void) { } EMIOS.MCR.B.GPRE= 63; EMIOS.MCR.B.ETB = 0; EMIOS.MCR.B.GPREN = 1; EMIOS.MCR.B.GTBE = 1; EMIOS.MCR.B.FRZ = 1; /* /* /* /* /* Divide 64 MHz sysclk by 63+1 = 64 for 1MHz eMIOS clk*/ External time base is disabled; Ch 23 drives ctr bus A */ Enable eMIOS clock */ Enable global time base */ Enable stopping channels when in debug mode */ void initEMIOSch23(void) { /* EMIOS CH 23: Modulus Up Counter */ EMIOS.CH[23].CADR.R = 999; /* Period will be 999+1 = 1000 clocks (1 msec) */ /* Use one of the following two lines for mode (Note some MPC555x devices lack MCB)*/ /*EMIOS.CH[23].CCR.B.MODE = 0x50;*//* MPC551x, MPC563x: Mod Ctr Bufd (MCB) int clk */ EMIOS.CH[23].CCR.B.MODE = 0x10;*//* MPC555x: Modulus Counter (MC) */ EMIOS.CH[23].CCR.B.BSL = 0x3; /* Use internal counter */ EMIOS.CH[23].CCR.B.UCPRE=0; /* Set channel prescaler to divide by 1 */ EMIOS.CH[23].CCR.B.FREN = 1; /* Freeze channel counting when in debug mode */ EMIOS.CH[23].CCR.B.UCPREN = 1; /* Enable prescaler; uses default divide by 1 */ } void initEMIOSch0(void) { /* EMIOS CH 0: Output Pulse Width Modulation */ EMIOS.CH[0].CADR.R = 250; /* Leading edge when channel counter bus=250*/ EMIOS.CH[0].CBDR.R = 500; /* Trailing edge when channel counter bus=500*/ EMIOS.CH[0].CCR.B.BSL = 0x0; /* Use counter bus A (default) */ EMIOS.CH[0].CCR.B.EDPOL = 1; /* Polarity-leading edge sets output/trailing clears*/ /* Use one of the following two lines for mode (Some MPC555x devices lack OPWMB)*/ /* EMIOS.CH[0].CCR.B.MODE = 0x60;*/ /* MPC551x, MPC563x: Mode is OPWM Buffered */ EMIOS.CH[0].CCR.B.MODE = 0x20; /* MPC555x: Mode is OPWM */ /* Use one of the following 2 lines: */ /* SIU.PCR[32].R = 0x0600; */ /* MPC551x: Initialize pad for eMIOS chan. 0 output */ SIU.PCR[179].R = 0x0E00; /* MPC555x: Initialize pad for eMIOS chan. 0 output */ } Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 153 void initEMIOSch1(void) { /* EMIOS CH 1: Output Pulse Width Modulation */ EMIOS.CH[1].CADR.R = 500; /* Leading edge when channel counter bus=500*/ EMIOS.CH[1].CBDR.R = 999; /* Trailing edge when channel's counter bus=999*/ EMIOS.CH[1].CCR.B.BSL = 0x0; /* Use counter bus A (default) */ EMIOS.CH[1].CCR.B.EDPOL = 1; /*Polarity-leading edge sets output/trailing clears*/ /* Use one of the following two lines for mode (Some MPC555x devices lack OPWMB) */ /* EMIOS.CH[1].CCR.B.MODE = 0x60;*/ /* MPC551x, MPC563x: Mode is OPWM Buffered */ EMIOS.CH[1].CCR.B.MODE = 0x20; /* MPC555x: Mode is OPWM */ /* Use one of the following 2 lines: */ /* SIU.PCR[33].R = 0x0600; */ /* MPC551x: Initialize pad for eMIOS chan. 1 output */ SIU.PCR[180].R = 0x0E00; /* MPC555x: Initialize pad for eMIOS chan. 1 output */ } void main (void) { volatile uint32_t i = 0; /* Dummy idle counter */ } initSysclk(); /* Set sysclk = 50MHz running from PLL */ initEMIOS(); /* Initialize eMIOS to provide 1 MHz clock to channels */ initEMIOSch23(); /* Initialize eMIOS channel 23 as modulus counter*/ initEMIOSch0(); /* Initialize eMIOS channel 0 as OPWM, using ch 23 as time base */ initEMIOSch1(); /* Initialize eMIOS channel 1 as OPWM, using ch 23 as time base */ while (1) {i++; } /* Wait forever */ 18.3.2 /* /* /* /* /* /* /* /* /* /* /* /* MPC56xxB/S (MPC56xxB shown with 8 MHz crystal) main.c - eMIOS OPWM example */ Description: eMIOS example using Modulus Counter and OPWM modes */ Rev 1.0 Sept 9 2004 S.Mihalik */ Rev 1.1 April 13 2006 S.M.- corrected GPRE to be div by 12 instead of 13*/ Rev 1.2 June 26 1006 S.M. - updated comments & made i volatile uint32_t */ Rev 1.3 July 19 2007 SM- Changes for MPC551x, 50 MHz sysclk, Mod Ctr data value*/ Rev 1.4 Aug 10 2007 SM - Changed to use sysclk of 64 MHz */ Rev 1.5 Jun 04 2008 SM - initSysclk changed for MPC5633M support */ Rev 1.6 May 22 2009 SM - modified for MPC56xxB/S */ Rev 1.7 Jun 24 2008 SM - simplified code */ Rev 1.8 Mar 14 2010 SM - modified initModesAndClock, updated header file */ Copyright Freescale Semiconductor, Inc. 2004–2010 All rights reserved. */ #include "MPC5604B_0M27V_0102.h" vuint32_t i = 0; /* Use proper include file */ /* Dummy idle counter */ void initModesAndClock(void) { ME.MER.R = 0x0000001D; /* Enable DRUN, RUN0, SAFE, RESET modes */ /* Initialize PLL before turning it on: */ /* Use 1 of the next 2 lines depending on crystal frequency: */ CGM.FMPLL_CR.R = 0x02400100; /* 8 MHz xtal: Set PLL0 to 64 MHz */ /*CGM.FMPLL_R = 0x12400100;*/ /* 40 MHz xtal: Set PLL0 to 64 MHz */ ME.RUN[0].R = 0x001F0074; /* RUN0 cfg: 16MHzIRCON,OSC0ON,PLL0ON,syclk=PLL */ ME.RUNPC[1].R = 0x00000010; /* Peri. Cfg. 1 settings: only run in RUN0 mode */ ME.PCTL[68].R = 0x01; /* MPC56xxB/S SIUL: select ME.RUNPC[1] */ ME.PCTL[72].R = 0x01; /* MPC56xxB/S EMIOS 0: select ME.RUNPC[1] */ /* Mode Transition to enter RUN0 mode: */ ME.MCTL.R = 0x40005AF0; /* Enter RUN0 Mode & Key */ ME.MCTL.R = 0x4000A50F; /* Enter RUN0 Mode & Inverted Key */ while (ME.GS.B.S_MTRANS) {} /* Wait for mode transition to complete */ /* Note: could wait here using timer and/or I_TC IRQ */ while(ME.GS.B.S_CURRENTMODE != 4) {} /* Verify RUN0 is the current mode */ } void initPeriClkGen(void) { CGM.SC_DC[2].R = 0x80; } /* MPC56xxB: Enable peri set 3 sysclk divided by 1*/ void disableWatchdog(void) { Qorivva Simple Cookbook, Rev. 4 154 Freescale Semiconductor } SWT.SR.R = 0x0000c520; SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; /* Write keys to clear soft lock bit */ /* Clear watchdog enable (WEN) */ void initEMIOS_0(void) { } EMIOS_0.MCR.B.GPRE= 63; /* Divide 64 MHz sysclk by 63+1 = 64 for 1MHz eMIOS clk*/ EMIOS_0.MCR.B.GPREN = 1;/* Enable eMIOS clock */ EMIOS_0.MCR.B.GTBE = 1; /* Enable global time base */ EMIOS_0.MCR.B.FRZ = 1; /* Enable stopping channels when in debug mode */ void initEMIOS_0ch23(void) { /* EMIOS 0 CH 23: Modulus Up Counter */ EMIOS_0.CH[23].CADR.R = 999; /* Period will be 999+1 = 1000 clocks (1 msec)*/ EMIOS_0.CH[23].CCR.B.MODE = 0x50; /* Modulus Counter Buffered (MCB) */ EMIOS_0.CH[23].CCR.B.BSL = 0x3; /* Use internal counter */ EMIOS_0.CH[23].CCR.B.UCPRE=0; /* Set channel prescaler to divide by 1 */ EMIOS_0.CH[23].CCR.B.UCPEN = 1; /* Enable prescaler; uses default divide by 1*/ EMIOS_0.CH[23].CCR.B.FREN = 1; /* Freeze channel counting when in debug mode*/ } void initEMIOS_0ch21(void) { /* EMIOS 0 CH 21: Output Pulse Width Modulation*/ EMIOS_0.CH[21].CADR.R = 250; /* Leading edge when channel counter bus=250*/ EMIOS_0.CH[21].CBDR.R = 500; /* Trailing edge when channel counter bus=500*/ EMIOS_0.CH[21].CCR.B.BSL = 0x0; /* Use counter bus A (default) */ EMIOS_0.CH[21].CCR.B.EDPOL = 1; /* Polarity-leading edge sets output */ EMIOS_0.CH[21].CCR.B.MODE = 0x60; /* Mode is OPWM Buffered */ SIU.PCR[69].R = 0x0600; /* MPC56xxS: Assign EMIOS_0 ch 21 to pad */ } void initEMIOS_0ch22(void) { /* EMIOS 0 CH 22: Output Pulse Width Modulation*/ EMIOS_0.CH[22].CADR.R = 500; /* Leading edge when channel counter bus=500*/ EMIOS_0.CH[22].CBDR.R = 999; /* Trailing edge when channel's counter bus=999*/ EMIOS_0.CH[22].CCR.B.BSL = 0x0; /* Use counter bus A (default) */ EMIOS_0.CH[22].CCR.B.EDPOL = 1; /* Polarity-leading edge sets output*/ EMIOS_0.CH[22].CCR.B.MODE = 0x60; /* Mode is OPWM Buffered */ SIU.PCR[70].R = 0x0600; /* MPC56xxS: Assign EMIOS_0 ch 22 to pad */ } void main (void) { volatile uint32_t i = 0; /* Dummy idle counter */ } initModesAndClock(); /* Initialize mode entries and system clock */ initPeriClkGen(); /* Initialize peripheral clock generation for DSPIs */ disableWatchdog(); /* Disable watchdog */ initEMIOS_0(); /* Initialize eMIOS 0 to provide 1 MHz clock to channels */ initEMIOS_0ch23(); /* Initialize eMIOS 0 channel 23 as modulus counter*/ initEMIOS_0ch21(); /* Initialize eMIOS 0 channel 0 as OPWM, ch 23 as time base */ initEMIOS_0ch22(); /* Initialize eMIOS 0 channel 1 as OPWM, ch 23 as time base */ while (1) {i++; } /* Wait forever */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 155 19 eMIOS: PEC, OPWFM Functions 19.1 Description Task: Count the number of input pulses in a time window using an eMIOS channel in pulse edge counting (PEC) mode. Input pulses will be generated by an eMIOS channel in output pulse width and frequency modulation (OPWFM) mode. The time base used for the time window is based on eMIOS channel 23 in modulus counter, as in Section 18, “eMIOS: Modulus Counter, OPWM Functions.” Also per that example, the window open and close times used for the PEC channel are replicated by a separate channel in OPWM mode, for observation purposes. Exercise: Connect eMIOS channels 2 and 3 externally. Optionally, connect an oscilloscope to observe the pulses and their mirrored counting window, channel 2. Execute the code and verify four pulses were counted in the window. Then alter code to reduce the timing window 50% and verify the result. MPC5500 Crystal 8 MHz eMIOS FMPLL with 1.5x Multiplier eMIOS Channel 23 (Mod. Counter: counts b1000) eMIOS Prescaler (Divide by 12) EMIOS[0] eMIOS Channel 3 (PEC: 400 s window Window open at 200 Window close at 600) 1 MHz eMIOS internal clock 12 MHz sysclk eMIOS Channel 0 (OPWM: Rising edge at 200 Falling edge at 600) EMIOS[3] eMIOS Channel 2 (OPWFM: 10 kHz, 100 s pulse period) EMIOS[2] Figure 32. eMIOS PEC and OPWFM Example Table 73. Signals for eMIOS PEC and OPWFM Example MPC555x Family Signal Function Name SIU PCR No. EMIOS[0] EMIOS[0] EMIOS[2] EMIOS[3] Package Pin No. 496 BGA 416 BGA 324 BGA 208 BGA 179 AD17 AF15 AB10 T4 EMIOS[2] 181 P21 AC16 W12 N7 EMIOS[3] 182 R22 AD15 AA11 R6 Qorivva Simple Cookbook, Rev. 4 156 Freescale Semiconductor 19.2 Design Timing resources used will include: • sysclk: 12 MHz: assume 8 MHz crystal and use default 1.5 multiplier • eMIOS internal clock: Choose 1 MHz (requires prescaling sysclk by 12) • eMIOS channel 2: Generate a 10 kHz signal using OPWFM mode • eMIOS channel 3: Count input pulses using PEC mode; use eMIOS channel 23 for timebase • eMIOS channel 23: Initialize in modulus counter mode, counting 1000 eMIOS internal clocks • eMIOS channel 0: OPWM mode based on Time Bus A — the output pulse will have the same timing parameters as the PEC window of eMIOS channel 3 19.2.1 Steps and Pseudo Code Table 74. eMIOS Pulse Edge Counting (PEC) and Output Pulse Width and Frequency Modulation (OPWFM) Step initEMIOS init EMIOS ch 2 (10 kHz OPWFM) Relevant Bit Field Config eMIOS for 1 MHz internal clock: • Set global prescaler (GPRE+1) = div. by 12 • Use channel 23 for Time Base A, not external • Enable global prescaler and internal clock • Disable freezing counters during debug GPRE = 0xB ETB = 0 (default) GPREN = 1 FRZ = 0 Pseudo Code EMIOS_MCR = 0x0400 0B00 Set period = 100 eMIOS clocks (1 s each) EMIOS_CH[2]CBDR = 99 Set duty cycle = 20 eMIOS clocks (1 s each) EMIOS_CH[2]CADR = 19 Set up Channel Control: • Channel counter’s prescaler = 1 • Enable counter’s prescaler • Set polarity to be active high • Mode = OPWFM, next period update, flag at B UCPRE = 0 UCPREN = 1 EDPOL = 1 MODE = 0x19 Configure GPIO 181: • Pad assignment = EMIOS Ch • Pad output buffer enabled PA = 3 OBE = 1 init EMIOS ch 23 Set modulus counter to match after 1000 clks (Modulus Counter) (Match value used for modulus counter) Set up Channel Control: • Mode = modulus up counter, using eMIOS internal clk • Counter bus select = internal • Set channel prescale to divide by • Enable channel prescaler EMIOS_CH[2] CCR= 0x0200 0099 SIU_PCR[181] = 0x0E00 A = 999 MODE = 0x10 BSL = 3 UCPRE = 0 UCPREN = 1 EMIOS_CH[23]CADR = 999 EMIOS_CH[23]CCR = 0x0200 0610 Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 157 Table 74. eMIOS Pulse Edge Counting (PEC) and Output Pulse Width and Frequency Modulation (OPWFM) Step init EMIOS ch 3 (PEC) Relevant Bit Field Define end of pulse counting window = 650 Set up Channel Control: • Use counter bus A for defining window • Count single edge trigger for counting • Count falling edges • Use main clock for input filtering • Use two clock periods for input filter • MODE = PEC, continuous mode EMIOS_CH[3]CBDR = 650 BSL = 0 EDSEL = 0 EDPOL = 0 FCK = 1 IF = 1 MODE =0xA Define start of pulse counting window = 250 (note: writing to this register after mode is set) init EMIOS ch 0 (OPWM) EMIOS_CH[3]CCR = 0x000C 000A EMIOS_CH[3]CADR = 250 Configure GPIO 182 • Pad assignment = EMIOS Ch 3 • Enable input buffer PA = 3 IBE = 1 SIU_PCR[182] = 0x0E00 Set Channel’s A match data value = 250 A = 250 EMIOS_CH[0]CADR = 250 Set Channel’s B match data value = 650 B = 650 EMIOS_CH[0]CBDR = 650 Set up Channel Control: • BSL = Bus selected is counter bus A BSL = 0 (default) • EDPOL = leading edge sets; trailing clears EDPOL = 1 • Mode = Output PWM, use immediate update MODE = 0x20 Configure GPIO 179: • Pad assignment = EMIOS Ch 0 • Enable output buffer PA = 3 OBE = 1 Start timers Start eMIOS (and eTPU) timers counting GTBE = 1 Read number of pulses Wait for channel flag to indicate end of window wait for FLAG = 1 Write one to FLAG bit to clear it EMIOS_CH[0]CCR = 0x0000 00A0 SIU_PCR[179] = 0x0E00 Read number of pulses counted Clear PEC flag Pseudo Code EMIOS_MCR[GTBE] = 1 wait for EMIOS_CH[3]CCR[FLAG] = 1 NoOfPulses = EMIOS_CH[3]CCNTR FLAG = 1 EMIOS_CH[3]CCR[FLAG] = 1 Qorivva Simple Cookbook, Rev. 4 158 Freescale Semiconductor 19.2.2 Design Screenshots Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 159 19.3 Code /* main.c - eMIOS example focusing on PEC and OPWFM*/ /* Description: Configures 1MHz eMIOS clock, Ch2 OPWFM at 10kHz, */ /* Ch 23 as mod ctr of 1MHz in & counts to 1000, Ch 3 as PEC */ /* Rev 1.0 Apr 20 2006, S. Mihalik - Initial version */ /* Rev 1.1 Jul 16 2007 SM - Corrected OPWFM B, A values to 99, 19 from 100, 20 */ /* and OPWM period to 999 from 1000 */ /* Copyright Freescale Semiconductor, Inc. 2006 All rights reserved. */ /* Notes: */ /* 1. MMU not initialized; must be done by debug scripts or BAM */ /* 2. SRAM not initialized; must be done by debug scripts or in a crt0 type file */ #include "mpc5554.h" void initEMIOS(void) { } EMIOS.MCR.B.GPRE= EMIOS.MCR.B.ETB = EMIOS.MCR.B.GPREN EMIOS.MCR.B.FRZ = 0xB; 0; = 1; 0; /* /* /* /* eMIOS clk= sysclk/(GPRE+1)= 12 MHz/12= 1MHz */ Ext. time base is disabled; Ch 23 drives ctr bus A */ Enable eMIOS clock */ Disable freezing channel counters in debug mode */ void initEMIOSch2(void) { } /* EMIOS CH 2: Output Pulse Width & Freq Modulation*/ /* Provide 10kHz output (100usec period) */ /* Input clock is eMIOS clk of 1MHz (1 usec period)*/ EMIOS.CH[2].CBDR.R = 99; /* Period = 1 usec x (99+1) = 100 usec, 10kHz*/ EMIOS.CH[2].CADR.R = 19; /* Duty cycle = 1 usec x (19+1) = 20 usec (20%) */ EMIOS.CH[2].CCR.B.UCPRE = 0; /* Channel counter uses divide by (0+1) prescaler */ EMIOS.CH[2].CCR.B.UCPREN = 1; /* Channel counter's prescaler is loaded & enabled*/ EMIOS.CH[2].CCR.B.EDPOL = 1; /* Polarity is active high */ EMIOS.CH[2].CCR.B.MODE= 0x19; /* Mode= 0PWFM, next period update, flag on B match*/ SIU.PCR[181].B.PA = 3; /* Initialize pad for eMIOS channel. */ SIU.PCR[181].B.OBE = 1; /* Initialize pad AC16 for output */ void initEMIOSch23(void) { /* EMIOS CH 23: Modulus Up Counter: */ /* input = 1MHz eMIOS clock, counts to 1000 */ EMIOS.CH[23].CADR.R = 999; /* Counter period= (999+1) clks x 10 usec/clk= 1 msec*/ EMIOS.CH[23].CCR.B.MODE = 0x10; /* Mode is Modulus Counter, internal clock */ EMIOS.CH[23].CCR.B.BSL = 0x3; /* Use internal counter */ EMIOS.CH[23].CCR.B.UCPRE=0; /* Set channel prescaler to divide by 1 */ EMIOS.CH[23].CCR.B.UCPREN = 1; /* Enable prescaler; uses default divide by 1 */ } void initEMIOSch3(void) { /* /* /* /* /* /* /* /* /* /* /* /* EMIOS CH 3: Pulse Edge Counting, single shot */ Count pulses during 400 usec window */ Count window closes when counter bus=650*/ Use counter bus A which is eMIOS Ch 23 */ Edge Select- Single edge trigger (count) */ Input filter will use main clock */ Input filger uses 2 clock periods */ Mode is PEC, continuous */ Count window opens when counter bus=250*/ NOTE: write to CADR after MODE is set */ Initialize pad for eMIOS channel */ Initialize pad for input */ /* /* EMIOS.CH[0].CADR.R = 250; /* EMIOS.CH[0].CBDR.R = 650; /* EMIOS.CH[0].CCR.B.BSL = 0x0; /* EMIOS.CH[0].CCR.B.EDPOL = 1; /* EMIOS.CH[0].CCR.B.MODE = 0x20; /* SIU.PCR[179].B.PA = 3; /* SIU.PCR[179].B.OBE = 1; /* EMIOS CH 0: Output Pulse Width Modulation */ Mirror ch 3 PEC window for observation */ Leading edge occurs when counter bus = 250 */ Trailing edge occurs when counter bus = 650 */ Use counter bus A which is eMIOS Ch 23 */ Polarity-leading edge sets output */ Mode is OPWM */ Initialize pad for eMIOS channel */ Initialize pad for output */ EMIOS.CH[3].CBDR.R = 650; EMIOS.CH[3].CCR.B.BSL = 0x0; EMIOS.CH[3].CCR.B.EDSEL = 0; EMIOS.CH[3].CCR.B.FCK = 1; EMIOS.CH[3].CCR.B.IF = 1; EMIOS.CH[3].CCR.B.MODE = 0xA; EMIOS.CH[3].CADR.R = 250; } SIU.PCR[182].B.PA = 3; SIU.PCR[182].B.IBE = 1; void initEMIOSch0(void) { } Qorivva Simple Cookbook, Rev. 4 160 Freescale Semiconductor void main (void) { uint32_t i = 0; vuint32_t NoOfPulses = 0; initEMIOS(); initEMIOSch2(); initEMIOSch23(); initEMIOSch3(); initEMIOSch0(); /* /* /* /* /* /* Dummy idle counter */ /* Number of pulses counted in PEC function of eMIOS */ Init. Init. Init. Init. Init. eMIOS to provide 1 MHz clock to eMIOS channels */ eMIOS channel 2 for 10kHz OPWFM */ eMIOS channel 23 as 1K modulus counter*/ eMIOS channel 3 for PEC, using ch 23 as time base */ eMIOS channel 0 as OPWM to mirror ch 3 PEC window */ EMIOS.MCR.B.GTBE = 1; /* Start timers/counters by enabling global time base */ } while (EMIOS.CH[3].CSR.B.FLAG == 0) {} /* Wait for flag at end of window */ NoOfPulses = EMIOS.CH[3].CCNTR.R; /* Read number of pulses counted */ EMIOS.CH[3].CSR.B.FLAG = 1; /* Clear flag */ while (1) {i++; } /* Wait forever */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 161 20 eTPU: Set 1 PWM Function 20.1 Description Task: Using the existing Set 1 eTPU functions from the Freescale web site, build an image that loads them to eTPU RAM. Then assign and use the PWM function on eTPU A channel 5, initially at 1 kHz with 25% duty cycle, then update to 2 kHz with 60% duty cycle. This example uses the Freescale-provided eTPU utilities, eTPU code image, and eTPU application program interface for the PWM function. For further information, see application note AN2864, General C Functions for the eTPU. Exercise: If using the MPC555x evaluation board, connect eTPU A channel 5 to the speaker, LED, or oscilloscope. Step through code and observe PWM frequency change. Then alter code to configure eTPU A channel 2 for PWM at a frequency of your choice. MPC5500 Crystal 8 MHz eTPUA eTPU A Channel 5 TCR1 with prescaler of sysclk/2 divided by 32 FMPLL with 16x multiplier ETPUA5 PWM: initial values: Period: 1000 Hz Duty Cycle: 25% eTPU A Channel 2 64 MHz sysclk ETPUA2 TCR1 counts at 1 MHz (available to all eTPUA channels and optionally to other eTPUs and eMIOS channels) Figure 33. eTPU Set 1 PWM Function Example Table 75. Signals for eTPU Set 1 PWM Example MPC555x Family Signal Function Name SIU PCR No. eTPUA5 ETPUA5 eTPUA2 ETPUA2 Package Pin No. EVB 496 BGA 416 BGA 324 BGA 208 BGA XPC 563M 119 H9 L4 K4 M4 PJ9–6 116 M3 M3 K3 P2 PJ9–3 Qorivva Simple Cookbook, Rev. 4 162 Freescale Semiconductor 20.2 Design Timing resources used will include: • sysclk = 64MHz: assume 8 MHz crystal, and use code from Section 10, “PLL: Initializing System Clock (MPC551x, MPC55xx),” to generate. • eTPU TCR1 clock: Count at 1 MHz rate. 20.2.1 Steps and Pseudo Code Table 76. Initialization: eTPU Using Set 1 Function Example Relevant Bit Field or Structure Step init FMPLL Set sysclk = 64 (See Section 10, “PLL: Initializing System Clock (MPC551x, MPC55xx)”) Pseudo Code / Function FMPLL_SYNCR = 0x1608 0000 Wait for FMPLL_SYNSR [LOCK] = 1; Note for 40 MHz Crystal used on MPC555x — Replace FMPLL_SYNCR values: 0x1608 0000 with 0x4610 0000 0x1600 0000 with 0x4608 0000 FMPLL_SYNCR = 0x1600 0000 Also for MPC563x: FMPLL_ESYNCR1[CLKCFG] = 7 init eTPU Configure eTPU A for: • MISC not used • eTPUA Input filter clock divided by 8 • eTPUA Channel input filter uses 3 samples • eTPUA TCR1 = sysclk/2 prescaled by 32 • eTPUA TCR2 = sysclk/8 prescaled by 8 • eTPUB configurations as desired init eTPUA[5] PWM • • • • • • Configure Pad Channel = eTPUA[5] Priority = middle Frequency = 1000 Hz Duty cycle = 25% Timebase = TCR1 Timebase frequency = 1 MHz Configure Pad for eTPUA[5] output Pad Assignment = eAPUA[5] Output Buffer is enabled Open Drain is not enabled etpu_config_t fs_etpu_init – fs_etpu_pwm_init SIU_PCR[119] = 0x0E00 PA = 3 OBE = 1 ODE = 0 Start timers Start all eTPU timers and eMIOS timers – fs_timer_start Update eTPU[5] • • • • – fs_etpu_pwm_update Channel = eTPUA[5] Frequency = 2000 Hz Duty cycle = 60% Timebase = 1 MHz Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 163 20.2.2 Files Used For Example Below is a summary of the files used for this project. All are available from the Freescale website, except for main.c and main.h, which are listed in Section 20.3, “Code.” Note that the eTPU C compiler is not needed to make this example because the output files are available on the Freescale website at www.freescale.com/etpu. However, all the source files used to build the set 1 functions are available and may be useful for reference. Table 77. Files Used in Example Provider User Freescale Type “Application code” Filename main.c main.h Description Files written for this example Link and Make Files link file make file Link file and make file used for this example eTPU Set 1 Library etpu_pwm.c etpu_pwm.h etpu_pwm_auto.h Host application program interface for pwm function Header file for pwm function Parameters automatically generated by eTPU compiler for pwm function eTPU Utilities etpu_set1.h Code image and globals generated by eTPU compiler for all of set 1 functions etpu_util.c etpu_util.h etpu_struct.h Host utilities to initialize eTPU, copy code image into code RAM, etc. MPC5500 Headers mpc5554.h mpc5554_vars.h typedefs.h Headers for MPC5500 device Qorivva Simple Cookbook, Rev. 4 164 Freescale Semiconductor 20.2.3 Design Screenshot The screenshot below shows the output at the initial PWM setting of 1 kHz period and 25% duty cycle. Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 165 20.3 20.3.1 Code main.c /***************************************************************************** * FILE NAME: main.c COPYRIGHT (c) FREESCALE 2006 * * DESCRIPTION: All Rights Reserved * * Sample eTPU program based on gpio_example function. * * Original author: J. Loelinger, modified by K Terry, G Emerson, S Mihalik * * 0.7 S. Mihalik 10/Aug/07 Modified for sysclk = 64 MHz * * 0.8 S. Mihalik 12/May/08 Added options for 40 MHz crystal, MPC563m * ****************************************************************************/ /* Notes: */ /* 1. MMU not initialized; must be done by debug scripts or BAM */ /* 2. SRAM not initialized; must be done by debug scripts or in a crt0 type file */ /* Use one of the next two pairs of lines for header files */ #include "mpc563m.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ #include "mpc563m_vars.h" /* MPC563m specific variables */ #include "etpu_util.h" /* useful utility routines */ #include "etpu_set1.h" /* eTPU standard function set 1 */ #include "etpu_pwm.h" /* eTPU PWM API */ /* User written include files */ #include "main.h" /* include application specific defines. */ uint32_t *fs_free_param; /* pointer to the first free parameter */ void initSysclk (void) { /* MPC563x: Use the next line */ FMPLL.ESYNCR1.B.CLKCFG = 0X7; /* Change clk to PLL normal mode from crystal */ /* MPC555x including MPC563x: use the next 3 lines for either 8 or 40 MHz crystal */ FMPLL.SYNCR.R = 0x16080000; /* 8 MHz xtal: 0x16080000; 40MHz: 0x46100000 */ while (FMPLL.SYNSR.B.LOCK != 1) {}; /* Wait for FMPLL to LOCK */ FMPLL.SYNCR.R = 0x16000000; /* 8 MHz xtal: 0x16000000; 40MHz: 0x46080000 */ } main () { int32_t error_code; /* Returned value from etpu API functions */ initSysclk(); /* Initialize PLL to 64 MHz */ /* Initialize eTPU hardware */ fs_etpu_init ( my_etpu_config, (uint32_t *) etpu_code, sizeof (etpu_code), (uint32_t *) etpu_globals, sizeof (etpu_globals)); /* Initialize eTPU channel ETPU_A[5] */ error_code = fs_etpu_pwm_init (5, /* Channel ETPU_A[5] */ FS_ETPU_PRIORITY_MIDDLE, 1000, /* Frequency = 1000 Hz*/ 2500, /* Duty cycle = 2500/100 = 25% */ FS_ETPU_PWM_ACTIVEHIGH, FS_ETPU_TCR1, 1000000); /* Timebase (TCR1) freq is 1 MHz */ SIU.PCR[119].R = 0x0E00; /* Configure pad for signal ETPU_A[5] output */ fs_timer_start (); /* Enable all timebases */ error_code = fs_etpu_pwm_update (5, /* Channel ETPU_A[5] */ 2000, /* New frequency = 2kHz*/ 6000, /* New duty cycle = 6000/100= 60% */ 1000000); /* Timebase (TCR1) freq = 1 MHz */ } while(1); /* Wait forever */ Qorivva Simple Cookbook, Rev. 4 166 Freescale Semiconductor 20.3.2 main.h /* main.h based on gpio_example.h below */ /************************************************************************* * FILE NAME: $RCSfile: gpio_example.h,v $ COPYRIGHT (c) FREESCALE 2004 * * DESCRIPTION: All Rights Reserved * * This file contains prototypes and definitions for the sample MPC5500 * * program using the the eTPU GPIO function. * *========================================================================* * ORIGINAL AUTHOR: Jeff Loeliger (r12110) * * $Log: gpio_example.h,v $ * Revision 1.1 2004/12/08 11:45:09 r47354 * Updates as per QOM API rel_2_1 * *........................................................................* * 0.1 J. Loeliger 05/Sep/03 Initial version. * * 0.2 K Terry 29/Apr/04 mod'd for GPIO function test * * 0.3 Updated for new build structure. * * 0.4 G. Emerson 2/Nov/04 Added etpu_config_t definition * **************************************************************************/ /* Rev 15/Mar/06 S. Mihalik : modified for eTPU PWM example */ /* Rev 15/Mar/06 S. Mihalik : modified for eTPU PWM example */ /* Rev 16/Jul/07 S. Mihalik : modified for 50 MHz sysclk, 1 MHz TCR1 */ /* Rev 10/Aug/07 S. Mihalik: modified for 64 MHz sysclk, still 1 MHz TCR1 */ #include "etpu_util.h" struct etpu_config_t my_etpu_config = { FS_ETPU_MISC_DISABLE, /*MCR register*/ FS_ETPU_MISC, /*MISC value from eTPU compiler link file*/ /*Configure eTPU engine A*/ FS_ETPU_FILTER_CLOCK_DIV8 + FS_ETPU_CHAN_FILTER_3SAMPLE + FS_ETPU_ENTRY_TABLE, FS_ETPU_TCR2CTL_DIV8 + ( 7 << 16) + FS_ETPU_TCR1CTL_DIV2 + 31, 0, /*Configure eTPU engine A timebases*/ /*TCR2 prescaler of 8 (7+1)*/ /*TCR1 prescaler of 32 (31+1) applied to sysclk/2*/ /*Configure eTPU engine B*/ FS_ETPU_FILTER_CLOCK_DIV4 + FS_ETPU_CHAN_FILTER_3SAMPLE + FS_ETPU_ENTRY_TABLE, }; FS_ETPU_TCR2CTL_DIV8 + ( 7 << 16) + FS_ETPU_TCR1CTL_DIV2 + 3, 0 /*Configure eTPU engine B timebases*/ /*TCR2 prescaler of 8 (7+1)*/ /*TCR1 prescaler of 4 (3+1)*/ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 167 21 eQADC: Single Software Scan 21.1 Description Task: Convert the analog signal on analog channel 5. Use CFIFO 0 in single-scan software-triggered mode, and use converter ADC0. Send results to RFIFO 0. The system clock will be the default of 16 MHz (MPC551x) or 12 MHz (MPC555x). The ADC_CLK, which must not exceed the 6 MHz for maximum resolution, will assume here the default system clock, which is not faster than 16 MHz. This minimal example shows how to send configuration commands for ADC0, send conversion command, and read the result. It does not incorporate queues, DMA, interrupts, calibration, or time stamp, nor does it clear the last Command FIFO EOQ flag and Result FIFO drain flag. Full accuracy may not be possible because calibration is not implemented here. For information, see AN2989, Design, Accuracy, and Calibration of Analog to Digital Converters on the MPC5500 Family. Exercise: Jumper a voltage, such as from a variable resistor, to AN5 and read the converted result. MPC5500 eQADC CFIFO0 Push Register CFIFO0 Trigger Mode: Single Scan, SW Trigger AN0 • • • AN5 • • • AN39 Configuration Commands CFIFO0 Conversion Commands • • • 40:1 MUX • • • ADC0_CR ADC0_TSCR ADC0_TBCR ADC0_GCCR ADC0_OCCR ADC0 (BN0) RFIFO0 RFIFO0 Pop Register Figure 34. eQADC Single Software Scan Example Table 78. Signals for eQADC Single Software Scan Example MPC551x Family Signal Pin Name SIU PCR No. AN5 PA[5] 5 MPC555x Family Package Pin No. 144 QFP 176 QFP 208 BGA 4 4 D1 Function Name SIU PCR No. AN[5] – Package Pin No. EVB 496 BGA 416 BGA 324 BGA 208 BGA xPC 563M B9 A8 A9 A7 PJ7–6 Qorivva Simple Cookbook, Rev. 4 168 Freescale Semiconductor 21.2 Design For MPC551x devices, the default system clock is 16 MHz. To achieve an ADC clock not exceeding 6 MHz, the prescaler will be 16 MHz sysclk / 6 MHz = 8/3 = 2.67. Because we need an even integer value that keeps the ADCCLK under 6 MHz, we round up to a value of prescaler of 4. This provides an ADCCLK = 16 MHz / 4 = 4 MHz. (Note for the MPC5516 evaluation board: the variable resistor is hard-wired to AN0 that is on PA[0], so you can simply jumper this to AN5, which is on PA[5], and run this program.) On MPC555x devices using the default system clock of 12 MHz, an ADC prescaler of 4 provides an ADCCLK = 12 MHz / 4 = 3 MHz. On MPC563x devices using the default system clock of 8 MHz, an ADC prescaler of 4 provides an ADCCLK = 8 MHz / 4 = 2 MHz. Table 79. eQADC Single Software Scan Step init ADC0 Relevant Bit Fields Determine Control Reg. value for ADC0: • Enable ADC0 • Prescaler = 4 —>ADC0 Control Reg = 0x8001 Send one write configuration command to CFIFO0: • End of Queue (only sending one message here) • Select ADC0 (Buffer Number 0) • Configuration command is Write (not Read) • ADC Control Register value = 0x8001 • ADC Control Register address = 0x1 Send Conversion Command Pseudo Code – ADC0_EN=1 ADC0_CLK_PS = 1(div 4) EOQ = 1 BN = 0 R/W = 0 (write) ADC_REGISTER=0x8001 ADC_REG_ADDRESS=1 EQADC_CFPR[0] = 0x8080_0101 Trigger CFIFO0 using single scan SW mode (Send configuration command(s) to ADC0’s registers) MODE0=1, SSE0=1 EQADC_CFCR[0] = 0x0410 Wait for End of Queue Flag for CFIFO0 wait for EOQF0 = 1 wait EQADC_FISR[EOQ] = 1 Clear End of Queue Flag for CFIFO0 EOQF = 1 Send one conversion command to CFIFO0: • Convert Channel 5 • Use Result FIFO0 • Use ADC0 (BN0) • Format is unsigned • Set EOQ CHANNEL_NUMBER = 5 MESSAGE_TAG = 0 BN = 0 FMT = 0 EOQ = 1 Trigger CFIFO0 using single scan SW mode (Sends conversion command(s) to ADC 0) Read Result Wait for RFIFO0 Drain Flag to set MODE0 = 1, SSE0 = 1 wait for RFDF0 = 1 EQADC_FISR[EOQ] = 1 EQADC_CRPR[0] = 0x8000_0500 EQADC_CFCR[0] = 0x0410 Wait EQADC_FISR[RFDF]=1 Read result from Result FIFO Pop Register 0 read EQADC_RFPR[0] Clear flags for any subsequent use. (Note: Flags are cleared by writing a 1. Code here is for illustrative purposes, but actually causes all flags in the FISR register to clear because the compiler will read the current value from the register, OR in the “1”, and write back the new value. Therefore existing flags at 1 are cleared. The proper way to clear a flag is to write to the entire register. EQADC_FISR [RFDF, EOQF] = 1 Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 169 21.3 /* /* /* /* /* /* /* /* /* Code main.c: performs a simple ADC conversion of channel 5 using ADC0 */ Rev 1.0 Sept 13, 2004 S. Mihalik. */ Rev 1.1 Jul 18 2007 SM- Changed ADCCLK prescaler for faster MPC551x default */ sysclk, added result in millivolts, used channel 5, conversion in loop */ Rev 1.2 Jun 12 2008 SM- Moved initADC0 out of while loop */ Copyright Freescale, 2007. All Rights Reserved */ Notes: */ 1. MMU not initialized; must be done by debug scripts or BAM */ 2. SRAM not initialized; must be done by debug scripts or in a crt0 type file */ #include "mpc563m.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ static uint32_t Result = 0; static uint32_t ResultInMv = 0; /* ADC conversion result */ /* ADC conversion result in millivolts */ void initADC0(void) { EQADC.CFPR[0].R = 0x80801001; } /* Send CFIFO 0 a ADC0 configuration command */ /* enable ADC0 & sets prescaler= divide by 2*/ EQADC.CFCR[0].R = 0x0410; /* Trigger CFIFO 0 using Single Scan SW mode */ while (EQADC.FISR[0].B.EOQF !=1) {} /* Wait for End Of Queue flag */ EQADC.FISR[0].B.EOQF = 1; /* Clear End Of Queue flag */ void SendConvCmd (void) { EQADC.CFPR[0].R = 0x80000500; /* Conversion command: convert channel 5 */ /* with ADC0, set EOQ, and send result to RFIFO 0*/ EQADC.CFCR[0].R = 0x0410; /* Trigger CFIFO 0 using Single Scan SW mode */ } void ReadResult(void) { while (EQADC.FISR[0].B.RFDF != 1){} /* Wait for RFIFO 0's Drain Flag to set*/ Result = EQADC.RFPR[0].R; /* ADC result */ ResultInMv = (uint32_t)((5000*Result)/0x3FFC); /* ADC result in millivolts */ EQADC.FISR[0].B.RFDF = 1; /* Clear RFIFO 0's Drain Flag */ EQADC.FISR[0].B.EOQF = 1; /* Clear CFIFO's End of Queue flag */ } int main(void) { int i = 0; initADC0(); while (1) { SendConvCmd(); ReadResult(); i++; } } /* Dummy idle counter */ /* Enable ADC0 only on eQADC */ /* Send one conversion command */ /* Read result */ /* Wait forever */ Qorivva Simple Cookbook, Rev. 4 170 Freescale Semiconductor 22 ADC: Software Trigger, Continuous Scan 22.1 Description Task: Convert a few standard ADC channel inputs by starting a normal conversion which is software triggered. Instead of One Shot Mode, Scan Mode is used where “a sequential conversions of N channels specified in the NCMR registers is continuously performed.”1 Exercise: Connect an analog channel to the pot on the EVB. Jumper ATD’s VDD to 5 V, download program, and verify results. Add additional channels and connect to a known voltage. Use ANS7 on PC[7] pin for the Dashboard Cluster Demo’s AUX MOTOR pot. Add an injected channel using an external trigger of PIT or eMIOS. MPC56xxB/P/S Clock Generation Module Crystal OSC0 64 MHz sysclk ADC_0 Normal Conversion Mask Register (NCMR0)* Ch 0* Ch 1* A/D Ch 2* . . . etc. ANS0 ANS1 ANS2 . . . etc. Channel Data Registers (CDRx)* CDR 0* CDR 1* CDR 2* . . . etc., * ANS0, ANS1, ANS2, etc. on MPC56xxB/S use NCMR1and CDR 32, 33, 34, etc. and on MPC560xP use NCMR0 and CDR 0, 1, 2, etc. Figure 35. ADC Continuous SW Scan Example Simplified Block Diagram Table 80. MPC56xxB/P/S Signals for Continous SW Scan Example Signal MPC56xxB Family Port MPC56xxP Family SIU Pad Package Pin # Configuration & Selection 100 144 176 Registers QFP QFP BGA (values in hex) Port MPC56xxS Family SIU Pad Package Port Configuration Pin # & Selection 100 144 Registers QFP QFP (values in hex) SIU Pad Package Pin # Configuration & Selection 144 176 208 Registers QFP QFP BGA (values in hex) ANS0 B8 PCR24=2000 ANS0 39 53 R9 B7 PCR23=2400 AN0 29 43 C0 PCR30=2000 ANS0 72 88 T13 ANS1 B9 PCR25=2000 ANS1 38 52 T9 B8 PCR24=2400 AN1 31 47 C1 PCR31=2000 ANS1 71 87 T12 ANS2 B10 PCR26=2000 ANS2 40 54 P9 C1 PCR33=2400 AN2 28 41 C2 PCR32=2000 ANS2 70 86 R12 1. MPC5606S Microcontroller Reference Manual, Rev 3, section 28.3.1.3, “Scan Mode description.” Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 171 22.2 22.2.1 Design Mapping of ADC Channels to Analog Input Pins Table 81. MPC56xxB/P/S Mapping of ADC Channels to ADC Input Pins MPC56xxB ADC Channel Number 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 MPC56xxP ADC 0 Pin Function Name ANP[0] ANP[1] ANP[2] ANP[3] ANP[4] ANP[5] ANP[6] ANP[7] ANP[8] ANP[9] ANP[10] ANP[11] ANP[12] ANP[13] ANP[14] ANP[15] MPC56xxS ADC 1 Pin Pin Function Function Name Name Pin Function Name AN[0] AN[0] AN[1] AN[1] AN[2] AN[2] AN[3] AN[3] AN[4] AN[4] AN[5] AN[5] AN[6] AN[6] AN[7] AN[7] AN[8] AN[8] AN[9] AN[9] AN[10] AN[10] AN[11] (shared) AN[12] (shared) AN[13] (shared) AN[14] (shared) Temp. 1.2V rail 32 ANS[0] ANS[0] 33 ANS[1] ANS[1] 34 ANS[2] ANS[2] 35 ANS[3] ANS[3] 36 ANS[4] ANS[4] 37 ANS[5] ANS[5] 38 ANS[6] ANS[6] 39 ANS[7] ANS[7] 40 ANS[8] ANS[8] 41 ANS[9] ANS[9] 42 ANS[10] ANS[10] 43 ANS[11] ANS[11] 44 ANS[12] ANS[12] 45 ANS[13] ANS[13] 46 ANS[14] ANS[14] 47 ANS[15] ANS[15] 48:59 ANS[26:17] The following channels use multiplex selector signals, MA[0:2] 64 ANX[0], MA=0 ANS[10], MA=0 65 ANX[0], MA=1 ANS[10], MA=1 Qorivva Simple Cookbook, Rev. 4 172 Freescale Semiconductor Table 81. MPC56xxB/P/S Mapping of ADC Channels to ADC Input Pins 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 22.2.2 ANX[0], MA=2 ANX[0], MA=3 ANX[0], MA=4 ANX[0], MA=5 ANX[0], MA=6 ANX[0], MA=7 ANX[1], MA=0 ANX[1], MA=1 ANX[1], MA=2 ANX[1], MA=3 ANX[1], MA=4 ANX[1], MA=5 ANX[1], MA=6 ANX[1], MA=7 ANX[2], MA=0 ANX[2], MA=1 ANX[2], MA=2 ANX[2], MA=3 ANX[2], MA=4 ANX[2], MA=5 ANX[2], MA=6 ANX[2], MA=7 ANX[3], MA=0 ANX[3], MA=1 ANX[3], MA=2 ANX[3], MA=3 ANX[3], MA=4 ANX[3], MA=5 ANX[3], MA=6 ANX[3], MA=7 ANS[10], MA=2 ANS[10], MA=3 ANS[10], MA=4 ANS[10], MA=5 ANS[10], MA=6 ANS[10], MA=7 Mode Use Mode Transition is required for changing mode entry registers. Hence even enabling the crystal oscillator to be active in the current mode (for example default mode (DRUN)) requires enabling the crystal oscillator in DRUN mode configuration register (ME_DRUN_MC), then initiating a mode transition to the same DRUN mode. This example changes from DRUN mode to RUN0 mode. This minimal example simply polls a status bit to wait for the targeted mode transition to complete. However, the status bit could instead be enabled to generate an interrupt request (assuming the INTC is intialized beforehand). This would allow software to complete other intialization tasks instead of brute force polling of the status bit. It is normal to use a timer when waiting for a status bit to change. This example by default would have a watchdog timer expire if for some reason the mode transition never completes. One could also loop code on incrementing a software counter to some maximum value as a timeout. If a timeout was reached, then an error condition could be recorded in EEPROM or elsewhere. Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 173 Table 82. Mode Configurations Summary for MPC56xxB/P/S ADC Continuous SW Scan Example (modes are enabled in ME_ME Register) Settings Mode Mode Config. Register Value sysclk Selection DRUN ME_DRUN_MC 0x001F 0010 (default) RUN0 Memory Power Mode Clock Sources Mode Config. Register ME_RUN0_MC 0x001F 0074 16MHz IRC XOSC0 PLL0 PLL1 (MPC Data 56xxP/S Flash only) Code Flash Main Voltage Reg. I/O Power Down Ctrl 16 MHz IRC On Off Off Off Normal Normal On Off PLL0 On On On Off Nomral Normal On Off Other modes are not used in example. Peripherals also have configurations to gate clocks on and off, enabling low power. The following table summarizes the peripheral configurations used here. ME_RUNPC_1 is selected, so peripherals to be used require a non-zero value in their respective ME_PCTL register. Table 83. Peripheral Configurations for MPC56xxB/P/S ADC Continous SW Scan Example (low power modes are not used in example) PeriPeri. Enabled Modes pheral Config. Config. Register RUN3 RUN2 RUN1 RUN0 DRUN SAFE TEST RESET PC1 ME_ RUNPC_ 1 0 0 0 1 0 0 0 0 Peripherals Selecting Configuration Peripheral PCTL Reg. # ADC 0 SIUL (MPC56xxB/S only) 32 68 Other peripheral configurations are not used in example. Qorivva Simple Cookbook, Rev. 4 174 Freescale Semiconductor 22.2.3 Steps and Pseudo Code Table 84. MPC5606B, MPC56xxP, MPC56xxS Steps for ADC Continous SW Scan Example Pseudo Code Step Relevant Bit Fields MPC56xxB Init Modes and Clock Enable desired modes RUN0, DRUN=1 MPC56xxP MPC56xxS ME_ME = 0x0000 001D Initialize PLL0 dividers to provide 64 MHz for input crystal before mode configuration: • 8 MHz xtal: FMPLL[0]_CR=0x02400100 • 40 MHz xtal: FMPLL[0]_CR=0x12400100 (MPC56xxP & 8 MHz xtal requires changing default CMU_CSR value. See PLL example.) CGM_FMPLL_CR (MPC56xxB) CGM_FMPLL[0]_CR (MPC56xxS) = 0x0240 0100 (8 MHz crystal) or 0x1240 0100 (40 MHz crystal) Configure RUN0 Mode: • I/O Output power-down: no safe gating • Main Voltage regulator is on (default) • Data, code flash in normal mode (default) • PLL0 is switched on • Crystal oscilator is switched on • 16 MHz IRC is switched on (default) • Select PLL0 (system pll) as sysclk PDO=0 MVRON=1 DFLAON = 3, CFLAON = 3 PLL0ON=1 XOSC0ON=1 16MHz_IRCON=1 SYSCLK=0x4 ME_RUN0_MC = 0x001F 0074 Peri. Config. 1: run in RUN0 mode only RUN0=1 ME_RUN_PC1 = 0x0000 0010 Assign peripheral configuration to peripherals: • ADC 0: select ME_RUN_PC0 RUN_CFG = 1 • SIUL: select ME_RUN_PC0 (56xxB/S) RUN_CFG = 1 Initiate software mode transition to RUN0 mode • Mode and key TARGET_MODE= • Mode and inverted key RUN0 • Wait for mode transition to complete S_MTRANS NOTE: if transition does not complete, check status flags such as ME_GS[XOSC] • Verify desired target mode was entered Disable • Write keys to clear soft lock bit Watchdog • Clear watchdog enable bit ME_PCTL32 = 0x01 .ME_PCTL68 = 0x01 (56xxB/S only) ME_MCTL =0x4000 5AF0 ME_MCTL =0x4000 A50F wait for ME_GS[S_MTRANS] = 0 verify ME_GS[S_CURRENT_MODE] = RUN0 SWT_SR = 0x000 0C520 SWT_SR = 0x0000 D928 SWT_CR = 0x8000 010A WEN = 0 init Peri Clk Gen Initialize peripheral clockgeneration (See appendix: MPC56xxB/P/S Peripheral Clocks) • ADC (56xxB/S): peripehral set 3- sysclk/1 init Pads Init pads for converting ANS0, ANS1, ANS2 CGM_ SC_DC2 = 0x80 DE2=1 DIV2=0 - CGM_ SC_DC2 = 0x80 SIU_PCR24 = SIU_PCR23 = SIU_PCR30 = SIU_PCR25= SIU_PCR24= SIU_PCR31= SIU_PCR26= SIU_PCR25= SIU_PCR33= 0x2000 0x2400 0x2000 Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 175 Table 84. MPC5606B, MPC56xxP, MPC56xxS Steps for ADC Continous SW Scan Example (continued) Pseudo Code Step Relevant Bit Fields MPC56xxB init ADC0 Initialize ADC 0 module & start conversions: • Configure Analog Clock as sysclk/2 (PLL/2 = 32 MHz for this example) • Mode is Scan Mode (continuous) • Trigger: on chip not used • ADC in normal mode, not power down mode MPC56xxP MPC56xxS ADC_MCR = 0x2000 0000 (MPC56xxB) ADC0_MCR = 0x2000 0000 (MPC56xxS) ADCLKSEL = 0 MODE = 1 TRIGEN =0 PWDN = 0 Enable normal sampling of desired channels for MPC56xxP: ADC 0 (Input pins ANS0, ANS1, ANS2) CHAN0:2 = 1 MPC56xxS: CHAN32:34 =1 ADC_ NCMR1 = 0x0000 0007 ADC0_ NCMR0 = 0x0000 0007 ADC0_ NCMR1 = 0x0000 0007 INPLATCH = 1 INPCMP = 3 INPSAMP = 6 ADC_CTR1 = 0x0000 8606 ADC0_CTR0 = 0x00008606 ADC0_CTR1 = 0x00008606 Initialize conversion timings for 32 MHz ADCLK1 Trigger ADC Start Normal Conversions: • Start normal conversion (immediately) Loop: Wait for completion of first chain (Note: VALID flag clears when read) wait for EOC = 1, Wait for first ADC_ ISR[EOC] =1 Read results CDATA Result0 = Result0 = Result0 = ADC_CDR32[ ADC0_CDR0 ADC0_CDR32[ CDATA] [CDATA] CDATA] Result1 = Result1 = Result1 = ADC_CDR33[ ADC0_CDR1 ADC0_CDR33[ CDATA] [CDATA] CDATA] Result2 = Result2 = Result2 = ADC_CDR34[ ADC0_CDR2 ADC0_CDR34[ CDATA] [CDATA] CDATA] Convert resutls to mv 1 ADC_MCR [NSTART] = 1 (MPC56xxB) ADC0_MCR [NSTART] = 1 (MPC56xxS) NSTART = 1 Wait for first ADC0_ ISR[EOC] =1 Wait for first ADC0_ ISR[EOC] =1 ResultInMv0 = int16_t (5000*Result0 / 0x3FF) ResultInMv1 = int16_t (5000*Result1 / 0x3FF) ResultInMv2 = int16_t (5000*Result2 / 0x3FF) Per “Max AD_CLK frequency and related configuration settings, AD_CLK” table row for 32 MHz fmax.in reference manuals: MPC5604B/C Microcontroller Reference Manual Rev 2 Table 25-1, MPC5604P Microcontroller Reference Manual Rev 2 Table 23-2, and MPC5606S Microcontroller Reference Manual Rev 3 Table 28-2. Qorivva Simple Cookbook, Rev. 4 176 Freescale Semiconductor 22.3 /* /* /* /* /* Code (MPC56xxS shown) main.c - ADC_ADC_scan example for MPC56xxS */ Description: Converts inputs ANS0, ANS1 using scan mode (continuous) */ Rev 1 Oct 26 2009 S Mihalik - initial version */ Rev 1.1 Mar 15 2010 S Mihalik- simplified initModesAndClock, new header */ Copyright Freescale Semiconductor, Inc 2009, 2010. All rights reserved. */ #include "56xxS_0200.h" /* Use proper header file */ uint16_t Result[3]; /* ADC conversion results */ uint16_t ResultInMv[3]; /* ADC conversion results in mv */ void initModesAndClock(void) { ME.MER.R = 0x0000001D; /* Enable DRUN, RUN0, SAFE, RESET modes */ /* Initialize PLL before turning it on: */ CGM.FMPLL[0].CR.R = 0x02400100; /* 8 MHz xtal: Set PLL0 to 64 MHz */ ME.RUN[0].R = 0x001F0074; /* RUN0 cfg: 16MHzIRCON,OSC0ON,PLL0ON,syclk=PLL0*/ ME.RUNPC[1].R = 0x00000010; /* Peri. Cfg. 1 settings: only run in RUN0 mode */ ME.PCTL[32].R = 0x01; /* MPC56xxB/P/S ADC 0: select ME.RUNPC[1] */ ME.PCTL[68].R = 0x01; /* MPC56xxB/S SIU: select ME.RUNPC[1] */ /* Mode Transition to enter RUN0 mode: */ ME.MCTL.R = 0x40005AF0; /* Enter RUN0 Mode & Key */ ME.MCTL.R = 0x4000A50F; /* Enter RUN0 Mode & Inverted Key */ while (ME.GS.B.S_MTRANS) {} /* Wait for mode transition to complete */ /* Note: could wait here using timer and/or I_TC IRQ */ while(ME.GS.B.S_CURRENTMODE != 4) {} /* Verify RUN0 is the current mode */ } void disableWatchdog(void) { SWT.SR.R = 0x0000c520; /* Write keys to clear soft lock bit */ SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; /* Clear watchdog enable (WEN) */ } void initPeriClkGen(void) { CGM.SC_DC[2].R = 0x80; /* MPC56xxB/S: Enable peri set 3 sysclk divided by 1 */ } void main (void) { vuint32_t i = 0; /* Dummy idle counter */ initModesAndClock(); disableWatchdog(); initPeriClkGen(); } /* Initialize mode entries and system clock */ /* Disable watchdog */ /* Initialize peripheral clock generation for DSPIs */ SIU.PCR[30].R = 0x2000; SIU.PCR[31].R = 0x2000; SIU.PCR[32].R = 0x2000; /* MPC56xxS: Initialize PC[0] as ANS0 */ /* MPC56xxS: Initialize PC[1] as ANS1 */ /* MPC56xxS: Initialize PC[2] as ANS2 */ ADC_0.MCR.R = 0x20000000; ADC_0.NCMR[1].R = 0x00000007; ADC_0.CTR[1].R = 0x00008606; ADC_0.MCR.B.NSTART=1; /* /* /* /* Initialize ADC0 for scan mode */ Select ANS0:2 inputs for conversion */ Conversion times for 32MHz ADClock */ Trigger normal conversions for ADC0 */ while (1) { while (ADC_0.CDR[33].B.VALID != 1) {}; /* Wait for last scan to complete */ Result[0]= ADC_0.CDR[32].B.CDATA; /* Read ANS0 conversion result data */ Result[1]= ADC_0.CDR[33].B.CDATA; /* Read ANS1 conversion result data */ Result[2]= ADC_0.CDR[34].B.CDATA; /* Read ANS2 conversion result data */ ResultInMv[0] = (uint16_t) (5000*Result[0]/0x3FF); /* Converted result in mv */ ResultInMv[1] = (uint16_t) (5000*Result[1]/0x3FF); /* Converted result in mv */ ResultInMv[2] = (uint16_t) (5000*Result[2]/0x3FF); /* Converted result in mv */ i++; } Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 177 23 ADC - CTU: eMIOS Trigger (MPC560xB) 23.1 Description Task: While performing normal conversions on a few standard ADC channel inputs, use the Cross Triggering Unit (CTU) to trigger a conversion from an eMIOS channel event, where the channel is configured as Single Action Input Capture (SAIC). In this example, ADC inputs ANS1 and ANS2 are continuously scanned. EMIOS and the CTU are configured so an input signal on eMIOS channel 2 will cause an event that “cross triggers” a conversion for ANS0. To generate an input signal, eMIOS channel 3 is configured for OPWM. Both channels use eMIOS channel 23 configured as modulus counter for their time base. Exercise: Make the connections shown below. Connect ANS0 to a potentiometer or a known voltage. Connect ANS1:2 to other known voltages. Verify the eMIOS channel 2 injects an ADC command which reads the pot value. MPC56xxB Crystal OSC0 64 MHz sysclk Clock Generation Module ADC Normal Conversion Mask Register (NCMR) Ch 32 Ch 33 A/D Ch 34 . . . etc. ANS0 ANS1 ANS2 . . . etc. Channel Data Registers CDR 32 CDR 33 CDR 34 . . . etc., Cross trigger ADC ch 32 eMIOS0 EMIOS0[2] Chan. 2 (SAIC) CTU Chan. 2 FLAG (with DMA=1) Event Cfg Reg 2 (Selects ADC chan 32) Chan. 3 (OPWM) EMIOS0[3] Chan. 23 (ModCtr) Figure 36. ADC CTU Example Simplified Block Diagram Qorivva Simple Cookbook, Rev. 4 178 Freescale Semiconductor Table 85. MPC56xxB Signals for ADC CTU Example Signal 23.2 23.2.1 MPC56xxB Family Port SIU Pad Configuration Package Pin # & Selection Registers 100 144 176 (values in hex) QFP QFP BGA ANS0 B8 PCR24=2000 39 53 R9 ANS1 B9 PCR25=2000 38 52 T9 ANS2 B10 PCR26=2000 40 54 P9 eMIOS0[2] A2 PCR2=0503 5 9 F2 eMIOS0[3] A3 PCR3=0600 68 90 K15 Design Channel Mappings The following table provides the numbering used for analog input signal names, ADC channel numbers, and CTU channel numbers. Table 86. MPC56xxB Mapping of Analog Signals, ADC channel numbers and CTU trigger channel numbers (per MPC5607B Microcontroller Reference Manual, Rev. 2) Signal Name (per Table 2-3) ADC Channel # (per Fig. 23-1) CTU_EVTCFTRx[CHANNEL_VALUE] (per Table 30-6) ADC 0 ADC 1 ANP 0:15 0:15 0:15 0:15 — — — - ANS 0:15 32:47 — 16:31 ANS 16:27 48:59 — Not mapped — — — — ANX 0, MA 0:7 64:71 — 32:39 ANX 0, MA 0:7 72:79 — 40:47 ANX 0, MA 0:7 80:87 — 48:55 ANX 0, MA 0:7 88:95 — 56:63 Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 179 23.2.2 Mode Use Mode Transition is required for changing mode entry registers. Hence even enabling the crystal oscillator to be active in the current mode (for example default mode (DRUN)) requires enabling the crystal oscillator in DRUN mode configuration register (ME_DRUN_MC), then initiating a mode transition to the same DRUN mode. This example changes from DRUN mode to RUN0 mode. This minimal example simply polls a status bit to wait for the targeted mode transition to complete. However, the status bit could instead be enabled to generate an interrupt request (assuming the INTC is intialized beforehand). This would allow software to complete other intialization tasks instead of brute force polling of the status bit. It is normal to use a timer when waiting for a status bit to change. This example by default would have a watchdog timer expire if for some reason the mode transition never completes. One could also loop code on incrementing a software counter to some maximum value as a timeout. If a timeout was reached, then an error condition could be recorded in EEPROM or elsewhere. Table 87. Mode Configurations Summary for MPC56xxB ADC CTU Example (modes are enabled in ME_ME Register) Settings Mode Mode Config. Register Value sysclk Selection DRUN ME_DRUN_MC 0x001F 0010 (default) RUN0 Memory Power Mode Clock Sources Mode Config. Register ME_RUN0_MC 0x001F 0074 16MHz IRC XOSC0 PLL0 PLL1 (MPC Data 56xxP/S Flash only) Code Flash Main Voltage Reg. I/O Power Down Ctrl 16 MHz IRC On Off Off Off Normal Normal On Off PLL0 On On On Off Nomral Normal On Off Other modes are not used in example. Peripherals also have configurations to gate clocks on and off, enabling low power. The following table summarizes the peripheral configurations used here. ME_RUNPC_1 is selected, so peripherals to be used require a non-zero value in their respective ME_PCTL register. Table 88. Peripheral Configurations for MPC56xxB ADC CTU Example (low power modes are not used in example) PeriPeri. Enabled Modes pheral Config. Config. Register RUN3 RUN2 RUN1 RUN0 DRUN SAFE TEST RESET PC1 ME_ RUNPC_ 1 0 0 0 1 0 0 0 0 Peripherals Selecting Configuration Peripheral ADC 0 CTUL SIUL eMIOS PCTL Reg. # 32 57 68 72 Other peripheral configurations are not used in example. Qorivva Simple Cookbook, Rev. 4 180 Freescale Semiconductor 23.2.3 Steps and Pseudo Code Table 89. MPC5606B Steps for MPC56xxB ADC CTU Example Step Init Modes and Clock Relevant Bit Fields Enable desired modes RUN0, DRUN=1 Initialize PLL0 dividers to provide 64 MHz for input crystal before mode configuration, using progessive clock switching: • 8 MHz Crystal: FMPLL[0]_CR=0x02400100 Pseudo Code ME_ME = 0x0000 001D CGM_FMPLL_CR (MPC56xxB) = 0x0240 0100 (8 MHz crystal) Configure RUN0 Mode: • I/O Output power-down: no safe gating • Main Voltage regulator is on (default) • Data, code flash in normal mode (default) • PLL0 is switched on • Crystal oscilator is switched on • 16 MHz IRC is switched on (default) • Select PLL0 (system pll) as sysclk PDO=0 MVRON=1 DFLAON = 3, CFLAON = 3 PLL0ON=1 XOSC0ON=1 16MHz_IRCON=1 SYSCLK=0x4 ME_RUN0_MC = 0x001F 0074 Peri. Config. 1: run in RUN0 mode only RUN0=1 ME_RUN_PC1 = 0x0000 0010 Assign peripheral configuration to peripherals: • ADC 0: select ME_RUN_PC0 • CTUL: select ME_RUN_PC0 • SIUL: select ME_RUN_PC0 • eMIOS 0: select ME_RUN_PC0 RUN_CFG = 1 RUN_CFG = 1 RUN_CFG = 1 RUN_CFG = 1 Initiate software mode transition to RUN0 mode • Mode and key • Mode and inverted key • Wait for mode transition to complete NOTE: if transition does not complete, check status flags such as ME_GS[XOSC] • Verify desired target mode was entered Disable • Write keys to clear soft lock bit Watchdog • Clear watchdog enable bit TARGET_MODE= RUN0 S_MTRANS ME_PCTL32 = 0x01 ME_PCTL57 = 0x01 .ME_PCTL68 = 0x01 .ME_PCTL72 = 0x01 ME_MCTL =0x4000 5AF0 ME_MCTL =0x4000 A50F wait for ME_GS[S._MTRANS] = 0 verify ME_GS[S_CURRENT_MODE] = RUN0 WEN = 0 init Peri Clk Gen Initialize peripheral clockgeneration (See appendix: MPC56xxB/P/S Peripheral Clocks) DE2=1 • ADC, CTU, eMIOS: peri. set 3- sysclk/1 DIV2=0 init Pads Init pads: • eMIOS0[2] as input • eMIOS0[3] as output • ANS0 • ANS1 • ANS2 SWT_SR = 0x000 0C520 SWT_SR = 0x0000 D928 SWT_CR = 0x8000 010A CGM_ SC_DC0 = 0x0000 8000 SIU_PCR2 = 0x0503 SIU_PCR3 = 0x0600 SIU_PCR24 = 0x2000 SIU_PCR25 = 0x2000 SIU_PCR26 = 0x2000 Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 181 Table 89. MPC5606B Steps for MPC56xxB ADC CTU Example (continued) Step init ADC Relevant Bit Fields Initialize ADC module but do not start conversions: • Configure Analog Clock as sysclk/2 (PLL/2 = 32 MHz for this example) • Mode is Scan Mode (continuous) • Trigger: on chip not used • ADC in normal mode, not power down mode • Enable CTU • Do not start conversions yet Pseudo Code ADC_MCR = 0x2002 0000 ADCLKSEL = 0 MODE = 1 TRIGEN =0 PWDN = 0 CTUEN = 1 NSTART = 0 Enable normal sampling of desired channels for ADC 0 (Input pins ANS1, ANS2) CHAN33:34 =1 ADC_NCMR1 = 0x0000 0006 Initialize conversion timings for 32 MHz ADCLK1 INPLATCH = 1 INPCMP = 3 INPSAMP = 6 ADC_CTR1 = 0x0000 8606 init CTU Configure event on eMIOS ch. 2 to trigger ADC ch. 0 • Enable trigger mask TM = 1 • Select input pint ANS0 (ADC ch 32=CTU ch 16) CHANNEL_VALUE = 16 (0x10) init eMIOS eMIOS module: configure for 1 MHz internal clock (see Modulus Counter, OPWM example for details) — EMIOS_0_MCR = 0x3000 3F00 Channel 23: initialize as modulus counter: Count 1000 1 usec clocks -> 1 kHz period (see Modulus Counter, OPWM example for details) — EMIOS_0_CHAN[23]CADDR = 999 EMIOS_0_CHAN[23]CCR = 0x8202 0650 Channel 3: initialize as OPWMB: Use Channel 23 as time base (see Modulus Counter, OPWM example for details) — EMIOS_0_CHAN[3]CADDR = 250 EMIOS_0_CHAN[3]CBDDR = 500 EMIOS_0_CHAN[3]CCR = 0x0000 00E0 Channel 2: intialize as SAIC • Bus selected is ch 23 • Edge Selected is a single edge • Edge Polarity is trigger on rising edge • Mode is SAIC • Flag enables IRQ or DMA request • DMA selected on flag (NOTE: CTU request generated instead of DMA request when CTU is enabled for that eMIOS channel) Normal trigger Start Normal Conversions: • Start normal conversion (immediately) Cross trigger Start eMIOS module counting, which will generate PWM output for eMIOS channel 2 input 1 BSL = 0 EDSEL = 0 EDPOL = 1 MODE = 2 FEN = 1 DMA = 1 CTU_EVTCFGR[2] = 0x0000 8010 EMIOS_0_CHAN[2]CCR = 0x0102 0082 ADC_MCR [NSTART] = 1 NSTART = 1 GTPREN = 1 EMIOS_0_MCR[GPREN] = 1 Per “Max AD_CLK frequency and related configuration settings, AD_CLK” table row for 32 MHz fmax.in reference manual: MPC5604B/C Microcontroller Reference Manual Rev 2 Table 25-1. Qorivva Simple Cookbook, Rev. 4 182 Freescale Semiconductor 23.3 /* /* /* /* /* /* Code main.c - ADC CTU example for MPC56xxB */ Description: Convert inputs ANS1:2 using normal scan mode and */ use eMIOS channel in SAIC mode with the CTU to trigger ANS0 */ Nov 11 2009 S Mihalik - initial version */ Mar 14 2010 S Mihalik - simplified initModesAndClock, updated header file */ Copyright Freescale Semiconductor, Inc 2009, 2010. All rights reserved. */ #include "MPC5604B_0M27V_0102.h" /* Use proper header file */ uint16_t Result; /* Read converstion result from ADC input ANS0 */ void initModesAndClock(void) { ME.MER.R = 0x0000001D; /* Enable DRUN, RUN0, SAFE, RESET modes */ /* Initialize PLL before turning it on: */ CGM.FMPLL_CR.R = 0x02400100; /* 8 MHz xtal: Set PLL0 to 64 MHz */ ME.RUN[0].R = 0x001F0074; /* RUN0 cfg: 16MHzIRCON,OSC0ON,PLL0ON,syclk=PLL0 */ ME.RUNPC[1].R = 0x00000010; /* Peri. Cfg. 1 settings: only run in RUN0 mode */ ME.PCTL[32].R = 0x01; /* MPC56xxB ADC 0: select ME.RUNPC[1] */ ME.PCTL[57].R = 0x01; /* MPC56xxB CTUL: select ME.RUNPC[1] */ ME.PCTL[68].R = 0x01; /* MPC56xxB SIU: select ME.RUNPC[1] */ ME.PCTL[72].R = 0x01; /* MPC56xxB eMIOS 0: select ME.RUNPC[1] */ /* Mode Transition to enter RUN0 mode: */ ME.MCTL.R = 0x40005AF0; /* Enter RUN0 Mode & Key */ ME.MCTL.R = 0x4000A50F; /* Enter RUN0 Mode & Inverted Key */ while (ME.GS.B.S_MTRANS) {} /* Wait for mode transition to complete */ /* Note: could wait here using timer and/or I_TC IRQ */ while(ME.GS.B.S_CURRENTMODE != 4) {} /* Verify RUN0 is the current mode */ } void disableWatchdog(void) { SWT.SR.R = 0x0000c520; /* Write keys to clear soft lock bit */ SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; /* Clear watchdog enable (WEN) */ } void initPeriClkGen(void) { CGM.SC_DC[2].R = 0x80; /* MPC56xxB/S: Enable peri set 3 sysclk divided by 1 */ } void initPads (void) { SIU.PCR[2].R = 0x0503; /* MPC56xxB: Initialize PA[2] as eMIOS[2] input */ SIU.PCR[3].R = 0x0600; /* MPC56xxB: Initialize PA[3] as eMIOS[3] output */ SIU.PCR[24].R = 0x2000; /* MPC56xxB: Initialize PB[8] as ANS0 */ SIU.PCR[25].R = 0x2000; /* MPC56xxB: Initialize PB[9] as ANS1 */ SIU.PCR[26].R = 0x2000; /* MPC56xxB: Initialize PB[10] as ANS2 */ } void initADC(void) { ADC.MCR.R = 0x20020000; /* Initialize ADC */ ADC.NCMR[1].R = 0x00000006; /* Select ANS1:2 inputs for normal conversion */ ADC.CTR[1].R = 0x00008606; /* Conversion times for 32MHz ADClock */ } void initCTU(void) { CTU.EVTCFGR[2].R = 0x00008010; /* Config event on eMIOS Ch 2 to trig ANS[0] */ } void initEMIOS_0(void) { EMIOS_0.MCR.R = 0x30003F00; /* Initialize eMIOS module for 1 MHz clock */ EMIOS_0.CH[23].CADR.R = 999; /* Ch 32: period will be 999+1 = 1K clks (1msec)*/ EMIOS_0.CH[23].CCR.R=0x82020650; /* Ch 32: set mode as modulus counter */ EMIOS_0.CH[3].CADR.R = 250; /* Ch 3: Match "A" is 250 */ EMIOS_0.CH[3].CBDR.R = 500; /* Ch 3: Match "B" is 500 */ EMIOS_0.CH[3].CCR.R= 0x000000E0; /* Ch 3: Mode is OPWMB, time base = ch 23 */ EMIOS_0.CH[2].CCR.R= 0x01020082; /* Ch 2: Mode is SAIC, time base = ch 23 */ } Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 183 void main (void) { vuint32_t i = 0; } /* Dummy idle counter */ initModesAndClock(); /* Initialize mode entries and system clock */ disableWatchdog(); /* Disable watchdog */ initPeriClkGen(); /* Initialize peripheral clock generation for DSPIs */ initPads(); /* Initialize pads used in example */ initADC(); /* Init. ADC for normal conversions but don't start yet*/ initCTU(); /* Configure desired CTU event(s) */ initEMIOS_0(); /* Initialize eMIOS channels as counter, SAIC, OPWM */ ADC.MCR.B.NSTART=1; /* Trigger normal conversions for ADC0 */ EMIOS_0.MCR.B.GPREN= 1; /* Start eMIOS counters to generate cross trigger */ while (1) { while (ADC.CDR[33].B.VALID != 1) {}; /* Wait for cross trigger to complete */ Result = ADC.CDR[32].B.CDATA; /* Read ANS0 conversion result data */ i++; } Qorivva Simple Cookbook, Rev. 4 184 Freescale Semiconductor 24 DSPI: SPI to SPI 24.1 Description Task: Transmit data from a DSPI configured as master to a DSPI configured as slave. The data sent by the master is 0x1234, and the slave will return 0x5678. This example shows how to configure a Clock and Transfer Attributes Register (CTAR) and other necessary items, then send a single command. Interrupts and DMA are not used. Initialization and data is set up first before clearing the DSPI’s HALT bit, which immediately enables SPI operation. Setting the transmit command’s End Of Queue (EOQ) bit puts the DSPI in the stopped state at the end of that frame’s transmission. Clearing EOQ re-enables transmission if commands are present. Exercise: (Note: external connections shown below are not required on MPC555x and MPC563x because of code that connects DSPIs internally in the SIU.) Change the data being sent and received. Debug tip: verify that SCK is at the desired baud rate by setting master DSPI MCR[CONT_SCK] with the debugger. PCS0_C MPC5500/ MPC5600 Crystal 8 MHz Master DSPI (DSPI_C or DSPI_0) Push Register TxFIFO Shift Register Clocks and PLL RFIFO CTAR0 SOUT_C SIN_C SCK_C Pop Register 64 MHz sysclk SCK_D Slave DSPI (DSPI_D or DSPI_B or DSPI_1) Push Register TxFIFO Shift Register SOUT_D SIN_D RFIFO CTAR0 Pop Register SS_D Figure 37. DSPI Single SPI Transfer Example (Signals shown use DSPI_C and DSPI_D) Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 185 Table 90. Signals for Single SPI Transfer Example using DSPI_C and DSPI_D Signal MPC555x Family Function Name SIU PCR # PCS0_C PCSC[0] SS_D Package Pin # 496 BGA 416 BGA 324 BGA 208 BGA 110 M26 R23 L19 J13 PCSD[0] 106 R28 N26 J22 H16 SCK_C SCKC 109 N27 P24 K20 H14 SCK_D SCKD 98 N26 T25 M21 J15 SIN_C SINC 108 M27 N24 J20 G14 SOUT_D SOUTD 100 U24 U23 N19 — SOUT_C SOUTC 107 T28 P26 K22 H15 SIND 99 N24 P23 K19 H13 SIN_D Table 91. Signals for Single SPI Transfer Example using DSPI_C and DSPI_B Signal MPC551x Family Pin Name SIU PCR # PB5 MPC555x Family Package Pin # Function Name SIU PCR # 144 QFP 176 QFP 208 BGA 21 129 157 D8 PCSC[0] PD12 60 90 114 H14 SCK_C PB6 22 128 156 SCK_B PD13 61 89 SIN_C PB8 24 SOUT_B PD14 SOUT_C PCS0_C SS_B SIN_B Package Pin # EVB 496 BGA 416 BGA 324 BGA 208 BGA xPC 563M 110 M26 R23 L19 J13 PJ2–9 PCSB[0] 105 R27 N25 J21 G16 PJ1–10 A9 SCKC 109 N27 P24 K20 H14 PJ2–8 113 H15 SCKB 102 T27 P25 K21 J16 PJ1–9 126 152 C9 SINC 108 M27 N24 J20 G14 PJ2–7 62 88 110 J14 SOUTB 104 N28 N23 J19 G13 PJ1–8 PB7 23 127 153 B9 SOUTC 107 T28 P26 K22 H15 PJ1–12 PD15 63 87 107 K14 SINB 103 P28 M26 H22 G15 PJ1–7 Qorivva Simple Cookbook, Rev. 4 186 Freescale Semiconductor Table 92. MPC56xxB/P/S Signals for Single SPI Transfer Example using DSPI_0 and DSPI_1 Signal MPC56xxB Family Port MPC56xxP Family MPC56xxS Family SIU Pad Package Pin # Port SIU Pad Package Port SIU Pad Package Pin # Configuration Configuration Pin # Configuration & Selection & Selection & Selection 100 144 176 100 144 144 176 208 Registers Registers Registers QFP QFP BGA QFP QFP QFP QFP BGA (values in hex) (values in hex) (values in hex) PCS0_0 A15 PCR15=0604 27 40 R6 C[4] PCR36=0604 5 11 SS_1 E5 PCR69=0903 PSMI9=02 94 133 C6 A[5] PCR5=0503 8 SCK_0 A14 PCR14=0604 28 42 P6 C[5] PCR37=0604 SCK_1 E4 PCR68=0903 PSMI7=01 93 132 SIN_0 A12 PCR12=0103 31 SOUT_1 C5 PCR37=0604 47 61 R5 14 C[13] PCR43=0D03 PSMI14=00 57 73 N9 7 13 B[9] PCR25=0604 44 54 T6 D6 A[6] PCR6=0503 2 2 B[4] PCR20=0503 48 62 P8 45 T7 C[7] PCR39=0103 9 15 B[7] PCR23=0503 46 56 P7 91 130 A7 A[7] PCR7=0604 4 10 B[5] PCR21=0604 49 63 N8 SOUT_0 A13 PCR13=0604 30 44 R7 C[6] PCR38=0604 98 142 B[8] PCR24=0604 45 55 N7 SIN_1 92 131 B7 A[8] PCR8=0103 6 12 B[6] PCR22=0503 50 66 R7 C4 PCR36=0103 PSMI8=00 H[4] PCR103=0604 Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 187 24.2 Design Determining the timing clock and transfer attributes, as well as the SPI command itself, depends entirely on the connected device’s specification. In order to show an example of how to determine these items, specifications from an actual device, the MC33394, are used (except where noted). 24.2.1 Clock and Transfer Attribute Register Parameters A 64 MHz sysclk is used in this example, which has a period of 15.625 nanoseconds. To keep the example simple, 15 nanoseconds will be used in the timing parameter calculations that follow. The master SPI only needs one CTAR, since there is just one device for SPI communication. Slave SPIs must use CTAR0. The parameters below are based on the MC33394 data sheet, rev 2.5 11/2002, p. 15, pp. 21–22. Both master and slave SPIs will use the same CTAR values. Frame Size: The MC33394 uses 16 bits for SPI communication. Clock Polarity: The MC33394 uses SCK low in the inactive state. Clock Phase: The MC33394 captures data on the leading edge, which is the rising edge since clock polarity is set to make SCK low when inactive. LSB First: Although the MC33394 requires the least significant bit is sent first, the MSB will be sent first in this example to make it easier to decipher the logic analyzer trace shown in the end of this section. PCS to SCK Delay (tCSC): The MC33394 calls this parameter the “enable lead time.” The MC33394 minimum value is 105 ns. The formula for PCS to SCK delay in the MPC5554 is: tCSC = (system clock period) (PCSSCK prescaler of 1,3,5 or 7) (CSSCK delay scaler of 2, 4, 8, etc.) With a system clock frequency of 64 MHz, the system clock period is about 15 nsec. This example will use: PCSSCK (PCS to SCK delay prescaler) = 0 (for a prescaler value of 1) CSSCK (PCS to SCK delay scaler) = 7 (for a scaler value of 256) This gives a PCS to SCK delay of: tCSC = 15 ns 1 256 = 3.84 s After SCK Delay (tASC): This is the delay from the last edge of SCK to the negation of PCS. The MC33394 calls this parameter “enable lag time.” The MC33394 minimum value is 50 ns. The formula for after SCK delay in the MPC5554 is: tASC = (system clock period) (PASC prescaler of 1,3,5 or 7) (ASC delay scaler of 2, 4, 8, 16 etc.) With a system clock frequency of 64 MHz, the system clock period is about 15 ns. This example will use: PASC (After SCK delay prescaler) = 0 (for a prescaler value of 1) ASC (After SCK delay scaler) = 7 (for a prescaler value of 256) This gives an after SCK delay of: tASC = 15 ns 1 256 = 3.84 s Qorivva Simple Cookbook, Rev. 4 188 Freescale Semiconductor Delay after Transfer (tDT): This is the length of time between the negation of PCS on the current frame and the assertion of PCS for the next frame. This example will only focus on a single transfer, but the calculation is shown here in case of a different situation. The MC33394 has a minimum time, called “CS Negated Time,” of 500 ns. tDT = (system clock period) (PDT prescaler of 1, 3, 5 or 7) (DT delay scaler of 2, 4, 8, 16 etc.) With a system clock frequency of 64 MHz, the system clock period is about 15 ns. This example will use: PDT (Delay after Transfer prescaler) = 2 (for a prescaler value of 5) DT (Delay after Transfer scaler) = 2 (for a prescaler value of 8) This gives an after SCK delay of: tDT = 15 ns 5 8 = 600 ns Baud Rate (BR): The baud rate maximum for the MC33394 is 5 MHz, but we will use close to 100 kHz here. The formula for DSPI baud rate in the MPC5554 is: SCK Baud Rate = (fSYS / PBR prescaler) ((1+DBR)/ (BR scaler)) With a system clock frequency of 64 MHz, this example will use: DBR = 0 (double baud rate feature not used) PBR = 2 (for a prescaler of 5) BR = 7 (for a scaler of 128) Hence the baud rate is SCK Baud Rate = (64 MHz / 5) ((1+0) / 128) = 100 kHz (10 s SCK period) 24.2.2 SPI Command The SPI command will be written to the SPI’s push register, which will then automatically fall through the FIFO. The fields are listed below, again based on the MC33394 data sheet. Continuous Peripheral Chip Select: Continuous PCS is not used and is inactive between transfers. Clock and Transfer Register: CTAR0 is required for the slave DSPI. The master will use CTAR0 also. End of Queue: Only one transfer is used in this example, so the EOQ bit will be set. Peripheral Chip Selects used: PCS0 will be connected from the master to the SS of the slave DSPI. Transfer Data: The slave DSPI (DSPI_D or DSPI_B) will have data 0x1234 in its shift register to respond to a transfer from the master DSPI (DSPI_C). The master will transmit 0x5678 for data. 24.2.3 Pad Slew Rate Control versus SPI Baud Rate Pads must be driven harder for faster signals. This example’s baud rate of 100 kHz is relatively slow for a SPI. Faster baud rates require increasing the SRC field value in the respective pad configuration registers, SIU_PCR[SRC]. Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 189 24.2.4 Mode Use Summary (MPC56xxB/P/S only) Mode Transition is required for changing mode entry registers. Hence even enabling the crystal oscillator to be active in the default mode (DRUN) requires enabling the crystal oscillator in appropriate mode configuration register (ME_xxxx_MC) then initiating a mode transition. This example transitions from the default mode after reset (DRUN) to RUN0 mode. Table 93. Mode Configurations for MPC56xxB/P/S DSPI SPI to SPI Example Modes are enabled in ME_ME Register. Settings Memory Power Mode Clock Sources Mode Mode Config. Register Mode Config. Register Value sysclk Selection PLL1 16MHz XOSC0 PLL0 (MPC IRC 56xxP/S only) Data Flash Code Flash Main Voltage Reg. I/O Power Down Ctrl DRUN ME_DRUN_MC 0x001F 0010 16 MHz IRC (default) On Off Off Off Normal Normal On Off RUN0 On On On Off Normal Normal On Off ME_RUN0_MC 0x001F 007D PLL0 Other modes are not used in example Peripherals also have configurations to gate clocks on or off for different modes, enabling low power. The following table summarizes the peripheral configurations used in this example. Table 94. Peripheral Configurations for MPC56xxB/P/S DSPI SPI to SPI Example Low power modes are not used in example. PeriPeri. Enabled Modes pheral Config. Config. Register RUN3 RUN2 RUN1 RUN0 DRUN SAFE TEST RESET PC1 ME_ RUNPC_ 1 0 0 0 1 0 0 0 0 Peripherals Selecting Configuration Peripheral DSPI 0 DSPI 1 SIUL (MPC56xxB/S) PCTL Reg. # 4 5 68 Other peripheral configurations are not used in example Qorivva Simple Cookbook, Rev. 4 190 Freescale Semiconductor 24.2.5 Steps and Pseudo Code Table 95. Steps and Pseudo Code Table (shown using DSPI_C as Master, DSPI_D as Slave) Relevant Bit Fields Step Data Init. Pseudo Code MPC551x MPC555x Initialize data to be received on master SPI Initialize data to be received on slave SPI MPC56xxB/P/S RecDataMaster = 0 RecDataSlave = 0 init Enable desired modes DRUN=1, Modes RUN0 = 1 and Clock Initialize PLL0 dividers to provide 64 MHz for input crystal before mode configuration: (MPC • 8 MHz xtal: FMPLL[0]_CR=0x02400100 56xxPBS • 40 MHz xtal: FMPLL[0]_CR=0x12400100 only) (MPC56xxP & 8 MHz xtal requires changing default CMU_CSR value. See PLL example.) - ME_ME = 0x0000 001D - 8 MHz Crystal: CGM_ FMPLL[0]_CR =0x02400100 Configure RUN0 Mode: • I/O Output power-down: no safe gating • Main Voltage regulator is on (default) • Data, code flash in normal mode (default) • PLL0 is switched on • Crystal oscillator is switched on • 16 MHz IRC is switched on (default) • Select PLL0 (system pll) as sysclk PDO=0 MVRON=1 DFLAON, CFLAON= 3 PLL0ON=1 XOSC0ON=1 16MHz_IRCON=1 SYSCLK=0x4 - ME_ RUN0_MC = 0x001F 0070 MPC56xxB/S: • Peri. Config. 1: run in RUN0 mode only RUN0=1 ME_RUN_PC1 = 0000 0010 - MC_PCTL4 = ME_PCTL5 = ME_PCTL68 = 0x01 - ME_MCTL =0x4000 5AF0, =0x4000 A50F wait ME_GS [S_TRANS] = 0 verify 4 = ME_GS [CURRENTMODE] Assign peripheral config. to peripherals: • DSPI 0: select ME_RUN_PC1 • DSPI 1: select ME_RUN_PC1 • SIUL: select ME_RUN_PC1 (MPC56xxB/S) RUN_CFG = 1 Initiate software mode transition to RUN0 mode • Mode & key, then mode & inverted key • Wait for transition to complete TARGET_MODE = RUN0 S_TRANS • Verify current mode is RUN0 CURRENTMODE - (Optional step - allows not using external connections for MPC555x, MPC563x only) • MPC555x: Connect master DSPI C to slave DSPI D internally init Peri Clk Gen (MPC 56xxPBS) - MPC563x: SIU_ISEL2.R = 0x00A8A000 • MPC563x: Connect master DSPI C to slave DSPI D internally init Sysclk Initialize sysclk to 64 MHz, running from PLL MPC555x: SIU_DISR = 0x0000 C0FC See PLL Initialization example - - MPC56xxB/S: CGM_SC_DC1= 0x80 Initialize peripheral clock generation (See appendix: MPC56xxB/P/S Peripheral Clocks) - MPC56xxB/S: Enable Peri Set 2- sysclk div. 1 Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 191 Table 95. Steps and Pseudo Code Table (shown using DSPI_C as Master, DSPI_D as Slave) (continued) Relevant Bit Fields Step Disable Watchdog Disable watchdog by writing keys to Status Register, then clearing WEN (MPC56xxBPS only) init DSPI_C or DSPI_0 (master) Initialize DSPI as master • SPI is master • SCK will not be continuous • DSPI is configured for SPI only • Debug Freeze does not halt transfers • Modified transfer format is not used • Peripheral chip select strobe is not used • Ignore receive FIFO overflows • Peripheral Chip Select 0 inactive state = high • Doze, module disable not used • Transmit FIFO not disabled • Receive FIFO not disabled • Transmit FIFO not cleared • Receive FIFO not cleared • Sample points not used • Halt state enabled (STOPPED) for initialization Configure CTAR0 for MC33394 • Frame Size = 16 bits • Clock polarity: low inactive state • Data is captured on leading edge • Most (not least) significant bit is first in frame • PCS to SCK delay = 3.84 usec (>105 ns required) • After SCK delay = 3.84 us (>50 ns required) • Delay after transfer = 800 ns (>500 ns required) • Double Baud Rate feature not used • Baud Rate = 78.125 kHz (<5 MHz required) Change DSPI from STOPPED to RUNNING Configure pads for Master (DSPI_C or DSPI_0): • SOUT_x output • SIN_x input • SCK_x output • PCSC0_x output Note: Faster baud rates would require increasing the pad’s slew rate control in SIU_PCR[SRC]. Pseudo Code MPC551x MPC555x MPC56xxB/P/S - See PLL Initialization example DSPIC_MCR = 0x8001 0001 DSPI0_MCR = 0x8001 0001 DSPIC_CTAR0 = 0x780A 7727 DSPI0_CTAR0 = 0x780A 7727 DSPIC_MCR = 0x8001 0000 DSPI0_MCR = 0x8001 0000 MSTR = 1 CONT_SCKE = 0 DCONF = 0 FRZ = 0 MTFE = 0 PCSSE = 0 ROOE = 0 PCSIS0 = 1 DOZE, MDIS = 0 DIS_TXF = 0 DIS_RXF = 0 CLR_TXF = 0 CLR_RXF = 0 SMPL_PT = 0 HALT = 1 FSIZE = 0xF CPOL = 0 CPHA = 0 LSBE = 0 PCSSCK = 0, CSSCK=7 PASC = 0, ASC = 7 PDT = 2, DT = 2 DBR = 0 PBR = 2, BR = 7 HALT = 0 SIU_PCR[23] = 0x0A00 SIU_PCR[24] = 0x0900 SIU_PCR[22] = 0x0A00 SIU_PCR[21] = 0x0A00 SIU_PCR[107] See table: = 0x0A00 MPC56xxB/P/S SIU_PCR[108] Signals = 0x0900 Connections Table SIU_PCR[109] for SIU_PCR = 0x0A00 Registers and SIU_PCR[110] Values = 0x0A00 Qorivva Simple Cookbook, Rev. 4 192 Freescale Semiconductor Table 95. Steps and Pseudo Code Table (shown using DSPI_C as Master, DSPI_D as Slave) (continued) Relevant Bit Fields Step init Enable as slave (other fields same as master MSTR=0 DSPI_D DSPI) or DSPI_1 Configure CTAR0 same as for master DSPI (Slave) Change DSPI from STOPPED to RUNNING HALT = 0 Configure pads for Slave (DSPI_B, DSPI_D, DSPI_1): • SCK_x input • SOUT_x output • PCS0/SS_x input Note: Faster baud rates would require increasing the pad’s slew rate control in SIU_PCR[SRC]. Transmit DSPI_C or DSPI 0 data Transmit a single SPI command from DSPI: • Continuous PCS disabled • Clock and Transfer Register 0 used • End of Queue is set (only one command sent). • Do not Clear SPI Transfer Count • PCS0=active, rest inactive • Transfer Data = 0x5678 MPC551x MPC555x MPC56xxB/P/S DSPID_MCR = 0x0001 0001 DSPI1_MCR = 0x0001 0001 DSPID_CTAR0 = 0x780A 7727 DSPI1_CTAR0 = 0x780A 7727 DSPID_MCR = 0x8001 0000 DSPI1_MCR = 0x8001 0000 SIU_PCR[42] = 0x0B00 SIU_PCR[44] = 0x0B00 SIU_PCR[43] = 0x0D00 SIU_PCR[45] = 0x1100 • SIN_x input Initialize response data from slave DSPI: Init • Data = 0x1234 DSPI_D or DSPI_1 Response Pseudo Code TxDATA=0x1234 CONT=0 CTAS=0 EOQ=1 CTCNT=0 PCS0=1, other PCSx=0 TxDATA=0x5678 SIU_PCR[98] = 0x0900 SIU_PCR[99] = 0x0900 SIU_PCR[100] = 0x0A00 SIU_PCR[106] = 0x0900 See table: MPC56xxB/P/S Signals for SIU_PCR & SIU_PSMI Registers and Values DSPID_PUSHR = 0x0000_1234 DSPI1_PUSHR = 0x0000_1234 DSPIC_PUSHR = 0x0801_5678 DSPI0_PUSHR = 0x0801_5678 Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 193 Table 95. Steps and Pseudo Code Table (shown using DSPI_C as Master, DSPI_D as Slave) (continued) Relevant Bit Fields Step Read data Wait for data to be received on slave SPI DSPI_D or DSPI_1 wait for RFDF=1 Read data on slave from master Clear flags by writing a “1” • Clear Receive FIFO Drain flag • Clear Transmit Complete flag Pseudo Code MPC551x wait for RDRF = 1 wait for DSPI1_SR[RDRF] =1 RecDataSlave = DSPID_POPR RecDataSlave = DSPI1_POPR DSPID_SR = 0x8002 0000 DSPI1_SR = 0x8002 0000 wait for DSPIC_SR[RDRF]=1 wait for DSPI0_SR[RDRF] =1 RecDataMaster = DSSPIC_POPR RecDataMaster = DSSPI0_POPR DSPIC_SR = 0x9002 0000 DSPI0_SR = 0x9002 0000 Read data on master from slave Clear flags by writing a “1” • Clear Receive FIFO Drain flag • Clear Transmit Complete flag • Clear End Of Queue flag 24.2.6 MPC56xxB/P/S wait for DSPID_SR[RDRF]=1 RFDF = 1 TCF = 1 Read data Wait for data to be received on master SPI DSPI_C or DSPI_0 MPC555x RFDF = 1 TCF = 1 EOQ=1 Design Screenshot The signals for this example are shown in the screenshot below for master DSPI_C and slave DSPI_D. The data is MSB first, and the frame size is 16 bits wide. Figure 38. DSPI Example Signals (Captured on MPC555x EVB) Qorivva Simple Cookbook, Rev. 4 194 Freescale Semiconductor 24.3 24.3.1 Code MPC551x (DSPI_C master, DSPI_B slave) /* main.c: performs a single transfer from DSPI_C to DSPI_B on MPC551x */ /* Rev 1.0 Sept 14 2004 S.Mihalik */ /* Rev 2.0 Jan 3 2007 S. Mihalik - Modified to use two SPIs */ /* Rev 2.1 July 20 2007 SM - Modified for MPC551x, changed sysclk (50 MHz) */ /* Rev 2.2 Aug 13 2007 SM - Modified for sysclk of 64 MHz & lenghened CSSCK, ASC*/ /* Rev 2.3 Jun 04 2008 SM - initSysclk changed for MPC5633M support */ /* Rev 2.4 Aug 14 2008 SM - Switched slave DSPI to DSPI_B, removed 555x lines */ /* Rev 2.5 Aug 18 2008 D McKenna-Kept DSPI_MCR[HALT] set during DSPI initialization*/ /* Rev 2.6 Aug 12 2009 SM - Changed PLL initial ERFD value, added 12MHz crystal */ /* Copyright Freescale Semiconductor, Inc. 2007 All rights reserved. */ #include "mpc5510.h" /* Use proper include file like mpc5510.h or mpc5554.h */ vuint32_t i = 0; /* Dummy idle counter */ uint16_t RecDataMaster = 0; /* Data recieved on master SPI */ uint16_t RecDataSlave = 0; /* Data received on slave SPI */ void initSysclk(void) { /* Initialize PLL and sysclk to 64 MHz */ /* Use appropriate code for crystal*/ FMPLL.ESYNCR2.R = 0x00000007; /* 8MHz xtal: ERFD to initial value of 7 */ FMPLL.ESYNCR1.R = 0xF0000020; /* 8MHz xtal: CLKCFG=PLL, EPREDIV=0, EMFD=0x20 */ /*FMPLL.ESYNCR2.R = 0x00000005; */ /* 12MHz xtal: ERFD to initial value of 5 */ /*FMPLL.ESYNCR1.R = 0xF0020030; */ /* 12MHz xtal: CLKCFG=PLL, EPREDIV=2, EMFD=0x30*/ CRP.CLKSRC.B.XOSCEN = 1; /* Enable external oscillator */ while (FMPLL.SYNSR.B.LOCK != 1) {}; /* Wait for PLL to LOCK */ FMPLL.ESYNCR2.R = 0x00000005; /* 8MHz xtal: ERFD change for 64 MHz sysclk */ /*FMPLL.ESYNCR2.R = 0x00000003; */ /* 12MHz xtal: ERFD change for 64 MHz sysclk */ SIU.SYSCLK.B.SYSCLKSEL = 2; /* Select PLL for sysclk */ } void initDSPI_C(void) { DSPI_C.MCR.R = 0x80010001; /* Configure DSPI_C as master */ DSPI_C.CTAR[0].R = 0x780A7727; /* Configure CTAR0 */ DSPI_C.MCR.B.HALT = 0x0; /* Exit HALT mode: go from STOPPED to RUNNING state*/ SIU.PCR[23].R = 0x0A00; /* MPC551x: Config pad as DSPI_C SOUT output */ SIU.PCR[24].R = 0x0900; /* MPC551x: Config pad as DSPI_C SIN input */ SIU.PCR[22].R = 0x0A00; /* MPC551x: Config pad as DSPI_C SCK output */ SIU.PCR[21].R = 0x0A00; /* MPC551x: Config pad as DSPI_C PCS0 output */ } void initDSPI_B(void) { DSPI_B.MCR.R = 0x00010001; /* Configure DSPI_B as slave */ DSPI_B.CTAR[0].R = 0x780A7727; /* Configure CTAR0 */ DSPI_B.MCR.B.HALT = 0x0; /* Exit HALT mode: go from STOPPED to RUNNING state*/ SIU.PCR[61].R = 0x0500; /* MPC551x: Config pad as DSPI_B SCK input */ SIU.PCR[63].R = 0x0500; /* MPC551x: Config pad as DSPI_B SIN input */ SIU.PCR[62].R = 0x0600; /* MPC551x: Config pad as DSPI_B SOUT output*/ SIU.PCR[60].R = 0x0500; /* MPC551x: Config pad as DSPI_B PCS0/SS input */ } void ReadDataDSPI_B(void) { while (DSPI_B.SR.B.RFDF != 1){} /* Wait for Receive FIFO Drain Flag = 1 */ RecDataSlave = DSPI_B.POPR.R; /* Read data received by slave SPI */ DSPI_B.SR.R = 0x80020000; /* Clear TCF, RDRF flags by writing 1 to them */ } void ReadDataDSPI_C(void) { while (DSPI_C.SR.B.RFDF != 1){} /* Wait for Receive FIFO Drain Flag = 1 */ RecDataMaster = DSPI_C.POPR.R; /* Read data received by master SPI */ DSPI_C.SR.R = 0x90020000; /* Clear TCF, RDRF, EOQ flags by writing 1 */ } int main(void) { initSysclk(); /* Set sysclk = 64MHz running from PLL */ initDSPI_C(); /* Initialize DSPI_C as master SPI and init CTAR0 */ initDSPI_B(); /* Initialize DSPI_B as Slave SPI and init CTAR0 */ DSPI_B.PUSHR.R = 0x00001234; /* Initialize slave DSPI_B's response to master */ DSPI_C.PUSHR.R = 0x08015678; /* Transmit data from master to slave SPI with EOQ */ ReadDataDSPI_B(); /* Read data on slave DSPI */ ReadDataDSPI_C(); /* Read data on master DSPI */ while (1) {i++; } /* Wait forever */ } Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 195 24.3.2 /* /* /* /* /* /* /* /* /* /* /* /* MPC555x (DSPI_C master, DSPI_D slave) main.c: performs a single transfer from DSPI_C to DSPI_D on MPC555x*/ Rev 1.0 Sept 14 2004 S.Mihalik */ Rev 2.0 Jan 3 2007 S. Mihalik - Modified to use two SPIs */ Rev 2.1 July 20 2007 SM - Modified for MPC551x, changed sysclk (50 MHz) */ Rev 2.2 Aug 13 2007 SM - Modified for sysclk of 64 MHz & lenghened CSSCK, ASC*/ Rev 2.3 Jun 04 2008 SM - initSysclk changed for MPC5633M support */ Rev 2.4 Aug 15 2008 SM - removed lines for MPC551x, MPC563x */ Rev 2.5 Aug 18 2008 D McKenna- Kept DSPI_MCR[HALT] set during initialization*/ Copyright Freescale Semiconductor, Inc. 2007 All rights reserved. */ Notes: */ 1. MMU not initialized; must be done by debug scripts or BAM */ 2. SRAM not initialized; must be done by debug scripts or in a crt0 type file */ #include "mpc5554.h" /* Use proper include file like mpc5510.h or mpc5554.h */ vuint32_t i = 0; uint16_t RecDataMaster = 0; uint16_t RecDataSlave = 0; /* Dummy idle counter */ /* Data recieved on master SPI */ /* Data received on slave SPI */ void initSysclk (void) { FMPLL.SYNCR.R = 0x16080000; /* 8 MHz xtal: 0x16080000; 40MHz: 0x46100000 */ while (FMPLL.SYNSR.B.LOCK != 1) {}; /* Wait for FMPLL to LOCK */ FMPLL.SYNCR.R = 0x16000000; /* 8 MHz xtal: 0x16000000; 40MHz: 0x46080000 */ } void initDSPI_C(void) { DSPI_C.MCR.R = 0x80010001; DSPI_C.CTAR[0].R = 0x780A7727; DSPI_C.MCR.B.HALT = 0x0; /* SIU.PCR[107].R = 0x0A00; SIU.PCR[108].R = 0x0900; SIU.PCR[109].R = 0x0A00; SIU.PCR[110].R = 0x0A00; } /* Configure DSPI_C as master */ /* Configure CTAR0 */ Exit HALT mode: go from STOPPED to RUNNING state*/ /* MPC555x: Config pad as DSPI_C SOUT output */ /* MPC555x: Config pad as DSPI_C SIN input */ /* MPC555x: Config pad as DSPI_C SCK output */ /* MPC555x: Config pad as DSPI_C PCS0 output */ void initDSPI_D(void) { DSPI_D.MCR.R = 0x00010001; /* Configure DSPI_D as slave */ DSPI_D.CTAR[0].R = 0x780A7727; /* Configure CTAR0 */ DSPI_D.MCR.B.HALT = 0x0; /* Exit HALT mode: go from STOPPED SIU.PCR[98].R = 0x0900; /* MPC555x: Config pad as DSPI_D SIU.PCR[99].R = 0x0900; /* MPC555x: Config pad as DSPI_D SIU.PCR[100].R = 0x0A00; /* MPC555x: Config pad as DSPI_D SIU.PCR[106].R = 0x0900; /* MPC555x: Config pad as DSPI_D } to RUNNING state*/ SCK input */ SIN input */ SOUT output*/ PCS0/SS input */ void ReadDataDSPI_D(void) { while (DSPI_D.SR.B.RFDF != 1){} RecDataSlave = DSPI_D.POPR.R; DSPI_D.SR.R = 0x80020000; } /* Wait for Receive FIFO Drain Flag = 1 */ /* Read data received by slave SPI */ /* Clear TCF, RDRF flags by writing 1 to them */ void ReadDataDSPI_C(void) { while (DSPI_C.SR.B.RFDF != 1){} RecDataMaster = DSPI_C.POPR.R; DSPI_C.SR.R = 0x90020000; } /* Wait for Receive FIFO Drain Flag = 1 */ /* Read data received by master SPI */ /* Clear TCF, RDRF, EOQ flags by writing 1 */ int main(void) { SIU.DISR.R = 0x0000C0FC; initSysclk(); initDSPI_C(); initDSPI_D(); DSPI_D.PUSHR.R = 0x00001234; DSPI_C.PUSHR.R = 0x08015678; ReadDataDSPI_D(); ReadDataDSPI_C(); while (1) {i++; } } /* /* /* /* /* /* /* /* /* MPC555x only: Connect DSPI_C, DSPI_D internally */ Set sysclk = 64MHz running from PLL */ Initialize DSPI_C as master SPI and init CTAR0 */ Initialize DSPI_D as Slave SPI and init CTAR0 */ Initialize slave DSPI_D's response to master */ Transmit data from master to slave SPI with EOQ */ Read data on slave DSPI */ Read data on master DSPI */ Wait forever */ Qorivva Simple Cookbook, Rev. 4 196 Freescale Semiconductor 24.3.3 MPC563x (DSPI_C master, DSPI_B slave) /* main.c: performs a single transfer from DSPI_C to DSPI_B */ /* Rev 1.0 Jun 2 2008 SM - Ported from AN2865 example Rev 2.2 for DSPI C, DSPI B */ /* and used POPR[RXDATA] for RecDataMaster, RecDataSlave */ /* Rev 1.1 Aug 15 2008 SM - Modified SIU.DISR line for internal DSPI connections */ /* Rev 1.2 Aug 18 2008 D McKenna- Kept DSPI_MCR[HALT] set during initialization*/ /* Copyright Freescale Semiconductor, Inc. 2007 All rights reserved. */ /* Notes: */ /* 1. MMU not initialized; must be done by debug scripts or BAM */ /* 2. SRAM not initialized; must be done by debug scripts or in a crt0 type file */ #include "mpc563m.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ vuint32_t i = 0; vuint32_t RecDataMaster = 0; vuint32_t RecDataSlave = 0; /* Dummy idle counter */ /* Data recieved on master SPI */ /* Data received on slave SPI */ void initSysclk (void) { /* MPC563x: Use the next line */ FMPLL.ESYNCR1.B.CLKCFG = 0X7; /* MPC563x: Change clk to PLL normal from crystal*/ FMPLL.SYNCR.R = 0x16080000; /* 8 MHz xtal: 0x16080000; 40MHz: 0x46100000 */ while (FMPLL.SYNSR.B.LOCK != 1) {}; /* Wait for FMPLL to LOCK */ FMPLL.SYNCR.R = 0x16000000; /* 8 MHz xtal: 0x16000000; 40MHz: 0x46080000 */ } void initDSPI_C(void) { DSPI_C.MCR.R = 0x80010001; /* Configure DSPI_C as master */ DSPI_C.CTAR[0].R = 0x780A7727; /* Configure CTAR0 */ DSPI_C.MCR.B.HALT = 0x0; /* Exit HALT mode: go from STOPPED SIU.PCR[107].R = 0x0A00; /* MPC555x: Config pad as DSPI_C SIU.PCR[108].R = 0x0900; /* MPC555x: Config pad as DSPI_C SIU.PCR[109].R = 0x0A00; /* MPC555x: Config pad as DSPI_C SIU.PCR[110].R = 0x0A00; /* MPC555x: Config pad as DSPI_C } void initDSPI_B(void) { DSPI_B.MCR.R = 0x00010001; /* Configure DSPI_B as slave */ DSPI_B.CTAR[0].R = 0x780A7727; /* Configure CTAR0 */ DSPI_B.MCR.B.HALT = 0x0; /* Exit HALT mode: go from STOPPED SIU.PCR[102].R = 0x0500; /* MPC555x: Config pad as DSPI_B SIU.PCR[103].R = 0x0500; /* MPC555x: Config pad as DSPI_B SIU.PCR[104].R = 0x0600; /* MPC555x: Config pad as DSPI_B SIU.PCR[105].R = 0x0500; /* MPC555x: Config pad as DSPI_B } to RUNNING state*/ SOUT output */ SIN input */ SCK output */ PCS0 output */ to RUNNING state*/ SCK input */ SIN input */ SOUT output*/ PCS0/SS input */ void ReadDataDSPI_B(void) { while (DSPI_B.SR.B.RFDF != 1){} /* Wait for Receive FIFO Drain Flag = 1 */ RecDataSlave = DSPI_B.POPR.B.RXDATA; /* Read data received by slave SPI */ DSPI_B.SR.R = 0x80020000; /* Clear TCF, RDRF flags by writing 1 to them */ } void ReadDataDSPI_C(void) { while (DSPI_C.SR.B.RFDF != 1){} /* Wait for Receive FIFO Drain Flag = 1 */ RecDataMaster = DSPI_C.POPR.B.RXDATA; /* Read data received by master SPI */ DSPI_C.SR.R = 0x90020000; /* Clear TCF, RDRF, EOQ flags by writing 1 */ } int main(void) { /* Optional: Use one of the next /* SIU.DISR.R = 0x0000C0FC; */ /* SIU.DISR.R = 0x00A8A000; /* initSysclk(); /* initDSPI_C(); /* initDSPI_B(); /* DSPI_B.PUSHR.R = 0x00001234; /* DSPI_C.PUSHR.R = 0x08015678; /* ReadDataDSPI_B(); /* ReadDataDSPI_C(); /* while (1) {i++; } /* } two lines for internal DSPI connections: */ MPC55xx except MPC563x: Connect DSPI_C, DSPI_D*/ MPC563x only: Connect DSPI_C, DSPI_B */ Set sysclk = 64MHz running from PLL */ Initialize DSPI_C as master SPI and init CTAR0 */ Initialize DSPI_B as Slave SPI and init CTAR0 */ Initialize slave DSPI_B's response to master */ Transmit data from master to slave SPI with EOQ */ Read data on slave DSPI */ Read data on master DSPI */ Wait forever */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 197 24.3.4 /* /* /* /* /* /* /* /* /* /* MPC56xxB/P/S (DSPI_0 master, DSPI_1 slave; MPC56xxP shown) main.c: performs a single transfer from DSPI_0 to DSPI_1 */ Rev 1.0 Sept 14 2004 S.Mihalik */ Rev 2.0 Jan 3 2007 S. Mihalik - Modified to use two SPIs */ Rev 2.1 July 20 2007 SM - Modified for MPC551x, changed sysclk (50 MHz) */ Rev 2.2 Aug 13 2007 SM - Modified for sysclk of 64 MHz & lenghened CSSCK, ASC*/ Rev 2.3 Mar 03 2009 SM - Modified for MPC56xxB/P/S */ Rev 2.4 May 22 2009 SM - Simplified code */ Rev 2.5 Jun 25 2009 SM - Simplified code */ Rev 2.6 Mar 14 2010 SM - modified initModesAndClock, updated header file */ Copyright Freescale Semiconductor, Inc. 2007–2010. All rights reserved. */ #include "Pictus_Header_v1_09.h" vuint32_t i = 0; uint16_t RecDataMaster = 0; uint16_t RecDataSlave = 0; /* Use proper include file */ /* Dummy idle counter */ /* Data recieved on master SPI */ /* Data received on slave SPI */ void initModesAndClks(void) { ME.MER.R = 0x0000001D; /* Enable DRUN, RUN0, SAFE, RESET modes */ /* Initialize XOSC, PLL before turning them on: */ /* Use 2 of the next 4 lines depending on crystal frequency: */ /*CGM.CMU_0_CSR.R = 0x000000004;*/ /* Monitor FXOSC > FIRC/4 (4MHz); no PLL monitor*/ /*CGM.FMPLL[0].CR.R = 0x02400100;*/ /* 8 MHz xtal: Set PLL0 to 64 MHz */ CGM.CMU_0_CSR.R = 0x000000000; /* Monitor FXOSC > FIRC/1 (16MHz); no PLL monitor*/ CGM.FMPLL[0].CR.R = 0x12400100; /* 40 MHz xtal: Set PLL0 to 64 MHz */ ME.RUN[0].R = 0x001F0074; /* RUN0 cfg: 16MHzIRCON,OSC0ON,PLL0ON,syclk=PLL */ ME.RUNPC[1].R = 0x00000010; /* Peri. Cfg. 1 settings: only run in RUN0 mode */ ME.PCTL[4].R = 0x01; /* MPC56xxB/P/S DSPI0: select ME.RUNPC[1] */ ME.PCTL[5].R = 0x01; /* MPC56xxB/P/S DSPI1: select ME.RUNPC[1] */ /*ME.PCTL[68].R = 0x01; */ /* MPC56xxB/S SIUL: select ME.RUNPC[1] */ /* Mode Transition to enter RUN0 mode: */ ME.MCTL.R = 0x40005AF0; /* Enter RUN0 Mode & Key */ ME.MCTL.R = 0x4000A50F; /* Enter RUN0 Mode & Inverted Key */ while (ME.GS.B.S_MTRANS) {} /* Wait for mode transition to complete */ /* Note: could wait here using timer and/or I_TC IRQ */ while(ME.GS.B.S_CURRENTMODE != 4) {} /* Verify RUN0 is the current mode */ } void initPeriClkGen(void) { /* Use the following code as required for MPC56xxB or MPC56xxS:*/ /*CGM.SC_DC[1].R = 0x80; */ /* MPC56xxB/S: Enable peri set 2 sysclk divided by 1*/ } void disableWatchdog(void) { SWT.SR.R = 0x0000c520; /* Write keys to clear soft lock bit */ SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; /* Clear watchdog enable (WEN) */ } void initDSPI_0(void) { DSPI_0.MCR.R = 0x80010001; /* Configure DSPI_0 DSPI_0.CTAR[0].R = 0x780A7727; /* Configure CTAR0 DSPI_0.MCR.B.HALT = 0x0; /* Exit HALT mode: go SIU.PCR[38].R = 0x0604; /* MPC56xxP: Config SIU.PCR[39].R = 0x0103; /* MPC56xxP: Config SIU.PCR[37].R = 0x0604; /* MPC56xxP: Config SIU.PCR[36].R = 0x0604; /* MPC56xxP: Config } as master */ */ from STOPPED to RUNNING state*/ pad as DSPI_0 SOUT output */ pad as DSPI_0 SIN input */ pad as DSPI_0 SCK output */ pad as DSPI_0 PCS0 output */ void initDSPI_1(void) { DSPI_1.MCR.R = 0x00010001; /* Configure DSPI_1 DSPI_1.CTAR[0].R = 0x780A7727; /* Configure CTAR0 DSPI_1.MCR.B.HALT = 0x0; /* Exit HALT mode: go SIU.PCR[6].R = 0x0503; /* MPC56xxP: Config SIU.PCR[8].R = 0x0103; /* MPC56xxP: Config SIU.PCR[7].R = 0x0604; /* MPC56xxP: Config SIU.PCR[5].R = 0x0503; /* MPC56xxP: Config } as slave */ */ from STOPPED to RUNNING state*/ pad as DSPI_1 SCK input */ pad as DSPI_1 SIN input */ pad as DSPI_1 SOUT output*/ pad as DSPI_1 PCS0/SS input */ Qorivva Simple Cookbook, Rev. 4 198 Freescale Semiconductor void ReadDataDSPI_1(void) { while (DSPI_1.SR.B.RFDF != 1){} RecDataSlave = DSPI_1.POPR.R; DSPI_1.SR.R = 0x80020000; } /* Wait for Receive FIFO Drain Flag = 1 */ /* Read data received by slave SPI */ /* Clear TCF, RDRF flags by writing 1 to them */ void ReadDataDSPI_0(void) { while (DSPI_0.SR.B.RFDF != 1){} RecDataMaster = DSPI_0.POPR.R; DSPI_0.SR.R = 0x90020000; } /* Wait for Receive FIFO Drain Flag = 1 */ /* Read data received by master SPI */ /* Clear TCF, RDRF, EOQ flags by writing 1 */ int main(void) { initModesAndClks(); initPeriClkGen(); disableWatchdog(); initDSPI_0(); initDSPI_1(); DSPI_1.PUSHR.R = 0x00001234; DSPI_0.PUSHR.R = 0x08015678; ReadDataDSPI_1(); ReadDataDSPI_0(); while (1) {i++; } } /* /* /* /* /* /* /* /* /* /* Initialize mode entries and system clock */ Initialize peripheral clock generation for DSPIs*/ Disable watchdog */ Initialize DSPI_0 as master SPI and init CTAR0 */ Initialize DSPI_1 as Slave SPI and init CTAR0 */ Initialize slave DSPI_1's response to master */ Transmit data from master to slave SPI with EOQ */ Read data on slave DSPI */ Read data on master DSPI */ Wait forever */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 199 25 FlexCAN Transmit and Receive 25.1 Description Task: Initialize two FlexCAN modules for 100 kHz bit time based on an 8 MHz crystal. Send a message from FlexCAN A (FlexCAN 0 on MPC56xxB/P/S). The modules will be connected externally in an open drain circuit, before the transceiver. The CAN reference will be based on the crystal, not the PLL. It is common for two CAN modules to be connected to the same transceiver. This allows more buffers to be available for a node. In this case, only one module is used for transmit and the other FlexCAN module’s transmit is not connected — otherwise the node might acknowledge itself. However, because there are not any other nodes connected in this example, we want the second CAN module to respond to the first. The connections needed are shown below. The transmit outputs must be configured as open drain and use a pullup resistor. Note this extra resistance on the transmit line may slow rise and fall times, which could limit the CAN frequency. Exercise: On a Evaluation Board (EVB), externally connect FlexCAN modules. EVBs typically have the pullup resistor. Be sure to use appropriate CAN_SEL jumpers on the EVB. Verify proper operation, then modify to send a second message. MPC5500 EVB +5V MPC5500 / MPC5600 Crystal 8 MHz Clocks and PLL 4.7K sysclk CANH CNTxA or FlexCAN_A or FlexCAN_0 Msg Buffer 0 Msg Buffer 1 ... Msg Buffer 63 CAN0TX CANL CNRxA or osc. Divide by (PRESDIV+1) MUX Time Quanta Freq. (SCK) XCVR CAN0RX CLK_SRC FlexCAN_C or FlexCAN 1 Msg Buffer 0 Msg Buffer 1 ... Msg Buffer 63 CNTxC or CAN1TX CNRxC or Divide by (PRESDIV+1) MUX Time Quanta Freq. (SCK) CAN1RX CLK_SRC Qorivva Simple Cookbook, Rev. 4 200 Freescale Semiconductor Table 96. MPC551x, MPC55xx Signals for FlexCAN Transmit and Receive Example MPC551x Family Signal Pin Name SIU PCR No. CNTxA PD0 CNTxC MPC555x Family Package Pin No. Function Name SIU PCR No. 144 QFP 176 QFP 208 BGA 48 104 128 D15 CNTXA PD4 52 100 124 E CNRxA PD1 49 103 127 CNRxC PD5 53 99 123 MPC56 Package Pin No. EVB 496 BGA 416 BGA 324 BGA 208 BGA xPC 563M 83 AF22 AD21 Y17 P12 PJ1–5 CNTXC 87 W24 V23 P19 K13 PJ1–6 D16 CNRXA 84 AG22 AE22 AA18 R12 PJ1–3 F13 CNRXC 88 Y26 W24 R20 L14 PJ1–4 Table 97. MPC56xxB/P/S Signals for FlexCAN Transmit and Receive Example (MPC56xxP: Safety Port is used for CAN1) Signal MPC56xxB Family Port CAN0TX MPC56xxP Family MPC56xxS Family SIU Pad Package Pin # Port SIU Pad Package Port SIU Pad Package Pin # Pin # Configuration & Configuration Configuration Selection & Selection & Selection 100 144 176 100 144 144 176 208 Registers Registers Registers QFP QFP BGA QFP QFP QFP QFP BGA (values in hex) (values in hex) (values in hex) B0 PCR16=0624 23 31 N3 B0 PCR16=0624 76 109 B0 PCR16=0624 CAN1TX C10 PCR42=0624 22 28 M3 A14 PCR14=0624 99 143 J7 PCR112=0A24 CAN0RX B1 PCR17=0100 24 32 N1 77 110 B1 PCR17=0500 PSMI0=00 CAN1RX C3 PCR35=0100 PSMI0=00 77 116 B11 A15 PCR15=0900 100 144 J6 PCR111=0900 PSMI1=02 B1 PCR17=0500 106 130 T15 - 60 A6 105 129 T14 - 59 B5 Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 201 Connection Notes: 1. MPC56XX EVB main board: needs CAN_SEL jumpers, J27, connecting pins 1–3 and 2–4. In addition to connecting both Rx pins together and both Tx pins together, at least one FlexCAN needs to be connected to the transceiver. This allows (1) the CAN bus to sync by allowing the Rx to see an idle bus and (2) the receiving FlexCAN to transmit an acknowledge from its Tx pin, through the transceiver, which routes it to the transmitting FlexCAN’s Rx pin. CANH_SEL (J27) jumpers are connected as shown and CANL_SEL (J29) jumpers are not connected. 2. MPC56XX EVB main board: I/O power (J4) should be jumpered to 5 V as shown so transceiver pullup gets 5 V. 3. MPC56XX main board with XPC560S: Trimmer jumper (J40) on main board connects Port B0 to an potentiometer hooked to 5V. Disconnect this jumper to allow Port B0 to drive it’s output properly. Qorivva Simple Cookbook, Rev. 4 202 Freescale Semiconductor 4. MPC56xxS EVB’s expansion card: make sure VDDE_B (J23)and VDDE_E (J25) are jumpered to 5V for the FlexCAN port pins used in this example. Transceiver’s need 5V on most EVBs. 5. MPC5510DEMO EVB needs CAN_EN jumpers (J9) connected by the CAN transceiver. Connect pins 1-2 and 3-4 to connect Tx and Rx to the transceiver as shown below. 6. MPC553x and MPC555x EVBs Tx and Rx pins need be connected to the CAN transceiver with jumpers. MPC5534EVB’s CAN_EN jumpers are shown below. Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 203 CAN_SEL jumpers 1, 2 installed on MPC5534EVB (default setting). These jumpers connect FlexCAN A Tx and Rx to transceiver. MPC5554EVB has jumpers located further to left. Debug Tips: 1. Observing a Repeating CAN Frame When debugging CAN, a simple test is only to connect Tx and Rx of one CAN together, and not to the transceiver. This direct connection allows the Rx pin to listen to the Tx without a transceiver, so that when a CAN frame starts to transmit then an error frame will not be generated. In addition, since there is no acknowledge from any receiving CAN module, the transmitting CAN module will keep repeating the transmitting of the frame, making it easy to examine on an oscilloscope or other tool. This example code uses an open drain on transmit pins, so if you are doing this test be sure either that an external pullup is connected on the board, or else disable the open drain in the transmitting CAN’s Pad Configuration Register (SIU_PCR). 2. Inadvertent Locking/Unlocking of Message Buffers Sometimes a message will not be received because a debugger window is reading the buffer’s Control and Status (CS), hence that buffer is locked and cannot receive a message. Conversely, a debugger window that constantly reads the TIMER or another message buffer may cause premature unlocking of a message buffer in your program. Qorivva Simple Cookbook, Rev. 4 204 Freescale Semiconductor 25.2 25.2.1 Design Timing Calculations These common guidelines are used in this example: • CAN bit rate period is typically subdivided into 12–20 time quanta units. • The sample point is normally chosen around 75%–80% through the bit rate period. • The remaining 20–25% will be the value for Phase_Seg2. • The value of Phase_Seg1 will be the same as Phase_Seg2. • The Sync_Seg is 1 time quanta. • Resync Jump Width (RJW+1) = Phase_Seg2 (if Phase_Seg2 < 4; otherwise (RJW +1) = 4. For this example and within the above guidelines, these are the values selected for CAN modules: number of time quanta units per bit rate period = 16 Sample point = 75%, which is 12 time quanta units into the 16 time quanta period Hence, Phase_Seg2 = (100% – 75%) 16 time quanta = 25% 16 time quanta = 4 time quanta; PSEG2 = 3 Phase_Seg1 = Phase_Seg2 = 4 time quanta; PSEG1 = 3 Prop_Seg = 16 – Phase_Seg1 – Phase_Seg2 – SYNCSEG = 16 – 4 – 4 – 1 = 7; PROPSEG = 6 Resync Jump Width (RJW + 1) = 4 Also for this example, the following applies for an 8 MHz crystal. (Note: If a 40 MHz crystal is used, the bit rate will increase five times.) fCANCLK = 8 MHz (EVB oscillator) Desired bit rate = 100 kHz Hence, ftq (time quanta frequency) = (16 time quanta/bit rate period) (100 K bit rate periods/sec) = 1.6 MHz Prescaler Value (PRESDIV + 1) = fCANCLK / ftq = 8 MHz / 1.6 MHz = 5 PRESDIV = 5 – 1 = 4 Table 98. Segments Within the Bit Time SYNCSEG PROP_SEG PHASE_SEG1 PHASE_SEG2 Number of Time Quanta 1 7 4 4 Register Bit Fields – PROPSEG + 1 PSEG1 + 1 PSEG 2 + 1 1 bit time = (16 time quanta) (1 sec / 1.6M time quanta) = 10 s Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 205 25.2.2 Message Buffers The message buffer structure is shown below, which is from Freescale’s MPC5554 header files version 1.2. Remember that buffers are in FlexCAN RAM, so they are random in value on power-up. Hence all buffers must all have their CODE field set inactive before negating the halt state. struct canbuf_t { union { vuint32_t R; struct { vuint32_t:4; vuint32_t CODE:4; vuint32_t:1; vuint32_t SRR:1; vuint32_t IDE:1; vuint32_t RTR:1; vuint32_t LENGTH:4; vuint32_t TIMESTAMP:16; } B; } CS; union { vuint32_t R; struct { vuint32_t:3; vuint32_t STD_ID:11; vuint32_t EXT_ID:18; } B; } ID; union { vuint8_t B[8]; vuint16_t H[4]; vuint32_t W[2]; vuint32_t R[2]; } DATA; /* /* /* /* Data Data Data Data buffer buffer buffer buffer in in in in Bytes (8 bits) */ Half-words (16 bits) */ words (32 bits) */ words (32 bits) */ } BUF[64]; . Qorivva Simple Cookbook, Rev. 4 206 Freescale Semiconductor 25.2.3 Mode Use Summary (MPC56xxB/P/S only) Mode Transition is required for changing mode entry registers. Hence even enabling the crystal oscillator to be active in the default mode (DRUN) requires enabling the crystal oscillator in appropriate mode configuration register (ME_xxxx_MC) then initiating a mode transition. This example transitions from the default mode after reset (DRUN) to RUN0 mode. Table 99. Mode Configurations for MPC56xxB/P/S FlexCAN Transmit and Receive Example Modes are enabled in ME_ME Register. Settings Memory Power Mode Clock Sources Mode Mode Config. Register Mode Config. Register Value sysclk Selection PLL1 16MHz XOSC0 PLL0 (MPC Data IRC 56xxP/S Flash only) Code Flash Main Voltage Reg. I/O Power Down Ctrl DRUN ME_DRUN_MC 0x001F 0010 16 MHz IRC (default) On Off Off Off Normal Normal On Off RUN0 On On On Off Normal Normal On Off ME_RUN0_MC 0x001F 007D PLL0 Other modes are not used in example Peripherals also have configurations to gate clocks on or off for different modes, enabling low power. The following table summarizes the peripheral configurations used in this example. Table 100. Peripheral Configurations for MPC56xxB/P/ S FlexCAN Transmit and Receive Example Low power modes are not used in example. PeriPeri. Enabled Modes pheral Config. Config. Register RUN3 RUN2 RUN1 RUN0 DRUN SAFE TEST RESET PC1 ME_ RUNPC_ 1 0 0 0 1 0 0 0 0 Peripherals Selecting Configuration Peripheral PCTL Reg. # FlexCAN0 (MPC56xxB/P/S) FlexCAN1 (MPC56xxB/S) SafetyPort (MPC56xxP use as FlexCAN) SIUL (MPC56xxB/S) 16 17 26 68 Other peripheral configurations are not used in example Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 207 25.2.4 Steps and Pseudo Code Table 101. Steps and Pseudo Code Table for FlexCAN Transmit and Receive Example Relevant Bit Fields Step Data Init Global variables for a CAN receive buffer Pseudo Code MPC551x MPC555x MPC56xxB/P/S RxCODE byte, RxID word, RxLENGTH byte, RxDATA 8 byte string, RxTIMESTAMP word, init Enable desired modes DRUN=1, Modes RUN0 = 1 and Clock Initialize PLL0 dividers to provide 64 MHz for input crystal before mode configuration: (MPC • 8 MHz xtal: FMPLL[0]_CR=0x02400100 56xxPBS • 40 MHz xtal: FMPLL[0]_CR=0x12400100 only) (MPC56xxP & 8 MHz xtal requires changing default CMU_CSR value. See PLL example.) — ME_ME = 0x0000 001D — 8 MHz Crystal: CGM_ FMPLL[0]_CR =0x02400100 Configure RUN0 Mode: • I/O Output power-down: no safe gating • Main Voltage regulator is on (default) • Data, code flash in normal mode (default) • PLL0 is switched on • Crystal oscillator is switched on • 16 MHz IRC is switched on (default) • Select PLL0 (system pll) as sysclk PDO=0 MVRON=1 DFLAON, CFLAON= 3 PLL0ON=1 XOSC0ON=1 16MHz_IRCON=1 SYSCLK=0x4 — ME_ RUN0_MC = 0x001F 0070 MPC56xxB/S: • Peri. Config.1: run in DRUN mode only — RUN0=1 ME_RUN_PC1 = 0000 0010 — .MC_PCTL16 = ME_PCTL17 = ME_PCTL96 = ME_PCTL68 = 0x01 - ME_MCTL =0x4000 5AF0, =0x4000 A50F wait ME_GS [S_TRANS] = 0 verify 4 = ME_GS [CURRENTMODE] — MPC56xxB/S: CGM_SC_DC1= 0x80 MPC56xxP: CGM_AC2_SC= 0x0400 0000 CGM_AC2_DC= 0x8000 0000 Assign peripheral configuration to peripherals: • FlexCAN 0: select ME_RUN_PC0 RUN_CFG = 1 • FlexCAN 1: select ME_RUN_PC0 RUN_CFG = 1 (56xxB/S) • Safety Port: select ME_RUN_PC0 (56xxP) RUN_CFG = 1 • SIUL: select ME_RUN_PC0 (56xxB/S) RUN_CFG = 1 Initiate software mode transition to RUN0 mode • Mode & key, then mode & inverted key • Wait for transition to complete TARGET_MODE = RUN0 S_TRANS • Verify current mode is RUN0 CURRENTMODE init Peri Clk Gen 56xxB/S Initialize peripheral clock generation (See appendix: MPC56xxB/P/S Peripheral Clocks) • MPC56xxB/S CANs: Peri Set 2- sysclk div. 1 • MPC56xxP Safety Port: Aux Clk 2- PLL0 div. 1 Qorivva Simple Cookbook, Rev. 4 208 Freescale Semiconductor Table 101. Steps and Pseudo Code Table for FlexCAN Transmit and Receive Example (continued) Relevant Bit Fields Step Pseudo Code MPC551x Disable Disable watchdog by writing keys to Status Watchdog Reg, then clearing WEN (MPC56xxBPS only) init FlexCAN C MPC56xxB/P/S — See PLL Initialization example CANC_MCR = 0x5000 003F CAN1_MCR = 0x5000 003F MPC56xxP only: 1F See other CAN CANC_CR = 0x04DB 0006 CAN1_CR = 0x04DB 0006 CODE = b0000 CANC_ MB0:63[CODE] = 0 CANC_MB4[IDE] = 0 CANC_MB4[ID] = 555 <<18 CANC_MB4[CODE] = 4 CAN1_ MB0:63[CODE] = 0 CAN1_MB4[IDE] = 0 CAN1_MB4[ID] = 555 <<18 CAN1_MB4[CODE]= 4 CANC_RXGMASK = 0x1FFFFFFF CAN1_RXGMASK = 0x1FFFFFFF Put FlexCAN module in freeze mode HALT= 1, FRZ= 1 When in freeze mode, enable all 64 buffers (32 message buffers for MPC56xxP) MAXMB = 63 or 31 (or Configure bit timing parameters, bit rate, FlexCAN arbitration (same as FlexCAN C or FlexCAN 1 or Safety 1) Port) Inactivate 32 or 64 Message Buffers, Initialize Message Buffer 4 for receive: • Extended ID is not used • Desired std. ID for incoming message is 555 (0x22B) • Set MB as RX EMPTY MPC555x IDE = 0 std. ID = 555 CODE = b0100 Initialize global acceptance mask: exact ID matches on all message buffers except 14, 15 MI=0x1FFF FFFF Initialize interrupt mask bits as needed (none used here) Configure pad as CNTXC, open drain, max slew rate Configure pad as CNRXC PA= 3, OBE= 1, ODE=1, SRC=3 PA= 3, IBE= 1 Negate HALT state, while enabling the maximum number of msg. buffers HALT=0, FRZ= 0 MAXMB = 63 or 31 – SIU_PCR[52 SIU_PCR[87] ] = 0x062C = 0x0E2C SIU_PCR[53 SIU_PCR[88] ] = 0x0500 = 0x0D03 See table: MPC56xxB/P/S Signals CANC_MCR = 0x0000 003F CAN1_MCR = 0x0000 003F or 1F Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 209 Table 101. Steps and Pseudo Code Table for FlexCAN Transmit and Receive Example (continued) Relevant Bit Fields Step init FlexCAN A Put FlexCAN module in freeze mode HALT=1, FRZ=1 When in freeze mode, enable all 64 buffers (32 message buffers for MPC56xxP) MAXMB = 63 (31 for MPC56xxP) (or Configure bit timing parameters FlexCAN0) Pseudo Code MPC551x MPC555x MPC56xxB/P/S CANA_MCR = 0x5000 003F CAN0_MCR = 0x5000 003F MPC56xxP only: 0x5000 001F CANA_CR = 0x04DB 0006 CAN0_CR = 0x04DB 0006 CANA_ MB0:63[CODE] = 0 CANA_ MB0[CODE] = 8 CAN0_ MB0:63[CODE]= 0 CAN0_ MB0[CODE] = 8 PROPSEG = 6 PSEG1 = 3 PSEG2 = 3 RJW = 3 Configure bit rate for 8 MHz OSC, 100 kHz bit CLK_SRC = 0 rate, and 16 time quanta PRESDIV = 4 Configure internal arbitration, others to default LBUF = 0 Inactivate all 64 message buffers (reduce to CODE = b0000 32 message buffers for MPC56xxP) CODE = b1000 Message Buffer 0: set as TX INACTIVE Initialize acceptance masks as needed (none used here) – Initialize interrupt mask bits as needed (none used here) – Configure pad as CNTXA, open drain, max slew rate Configure pad as CNRXA PA = 1, OBE =1, ODE=1, SRC=3 PA = 1, IBE = 1 Transmit Message Negate HALT state, while enabling the maximum number of message buffers HALT=0, FRZ=0 MAXMB = 63 or 31 (Assume CANA MB 0 is inactive) Use standard ID, not extended ID Standard ID = 555 (0x22B) IDE = 0 std. ID = 555 Frame is a data frame, not remote Tx request RTR = 0 Length of data is five bytes LENGTH = 5 Data is string “Hello” DATA = ‘Hello’ SRR = 1 for transmit; only req’d in exd’d frames] Activate MB to transmit a data frame SRR = 1 SIU_PCR[48] SIU_PCR[83] = 0x062C = 0x062C SIU_PCR[49] SIU_PCR[84] = 0x0500 = 0x0500 See table: MPC56xxB/P/S Signals CANA_MCR = 0x0000 003F CAN0_MCR = 0x0000 003F or 1F CANA_MB0[IDE] = 0 CANA_MB0[ID] = 555 <<18 CANA_MB0[RTR] = 0 CANA_MB0[LENGTH] =5 CANA_MB0[DATA] = ‘Hello’ CAN0_MB0[IDE]= 0 CAN0_MB0[ID] = 555 <<18 CAN0_MB0[RTR] = 0 CAN0_MB0[LENGTH ]=5 CAN0_MB0[DATA] = ‘Hello’ CAN0_MB0[SRR] = 1 CAN0_MB0[CODE] = 0xC CANA_MB0[SRR] = 1 CANA_MB0[CODE] = 0xC CODE = b1100 Qorivva Simple Cookbook, Rev. 4 210 Freescale Semiconductor Table 101. Steps and Pseudo Code Table for FlexCAN Transmit and Receive Example (continued) Relevant Bit Fields Step Receive Message Wait for CANC MB 4 flag to set wait for BUF04I = 1 Read CODE (activates internal buffer lock) CODE will be b0010 Read ID Read DATA Read TIMESTAMP Read TIMER to unlock buffer CODE = b0100 Clear CANC MB 4 flag by writing a 1 to it BUF04I = 1 Pseudo Code MPC551x MPC555x MPC56xxB/P/S wait for wait for CANC_ CANC_ IFLAG1 IFRL [BUF04I] = 1 [BUF04I] = 1 wait for CAN1_ IFRL [BUF04I] = 1 RxCODE = CANC_MB4[CODE] RxID = CANC_MB4[ID] RxDATA[ ] = CANC_MB4[DATA] RxTIMESTAMP = CANC_MB4 [TIMESTAMP] RxCODE = CAN1_MB4[CODE] RxID = CAN1_MB4[ID] RxDATA[ ] = CAN1_MB4[DATA] RxTIMESTAMP = CAN1_MB4 [TIMESTAMP] Read CANC_TIMER Read CAN1_TIMER CANC_ CANC_ IFLAG1 = IFRL = 0x0000 0010 0x0000 0010 CAN1_ IFRL = 0x0000 0010 Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 211 25.2.5 Design Results The pictures below show results for this design based on an 8 MHz crystal. • Oscilloscope screenshot of the start of the transmit output line, changing at the 100 kHz bit time rate • CAN tool screenshot of the transmitted message as seen on the CAN bus Qorivva Simple Cookbook, Rev. 4 212 Freescale Semiconductor 25.3 25.3.1 /* /* /* /* /* /* /* /* /* /* /* /* /* /* /* /* /* /* /* Code MPC551x, MPC555x main.c: FlexCAN program */ Description: Transmit one message from FlexCAN A buf. 0 to FlexCAN C buf. 4 */ Rev 0.1 Jan 16, 2006 S.Mihalik, Copyright Freescale, 2006. All Rights Reserved */ Rev 0.2 Jun 6 2006 SM - changed Flexcan A to C & enabled 64 msg buffers */ Rev 0.3 Jun 15 2006 SM - 1. Made globals uninitialized */ 2. receiveMsg function: read CANx_TIMER, removed setting buffer's CODE*/ 3. added idle loop code for smoother Nexus trace */ 4. modified for newer Freescale header files (r 16) */ Rev 0.4 Aug 11 2006 SM - Removed redundant CAN_A.MCR init */ Rev 0.5 Jan 31 2007 SM - Removed other redundant CAN_C.MCR init */ Rev 0.6 Mar 08 2007 SM - Correced init of MBs - cleared 64 MBs, instead of 63 */ Rev 0.7 Jul 20 2007 SM - Changes for MPC5510 */ Rev 0.8 May 15 2008 SM - Changes for new MPC5510 header file symbols */ Rev 1.0 Jul 10 2009 SM - Cleared CAN Msg Buf flag by writing to reg. not bit, */ increased Tx pads slew rate, changed RxCODE, RxLENGTH, dummy data types and */ init receiving CAN first to allow CAN bus sync time before receiving first msg*/ Notes: */ 1. MMU not initialized; must be done by debug scripts or BAM */ 2. SRAM not initialized; must be done by debug scripts or in a crt0 type file */ #include "mpc563m.h" /* Use proper include file such as mpc5510.h or mpc5554.h */ uint32_t RxCODE; uint32_t RxID; uint32_t RxLENGTH; uint8_t RxDATA[8]; uint32_t RxTIMESTAMP; /* /* /* /* /* Received Received Received Received Received message message message message message buffer code */ ID */ number of data bytes */ data string*/ time */ void initCAN_A (void) { uint8_t i; CAN_A.MCR.R = 0x5000003F; /* Put in Freeze Mode & enable all 64 msg buffers*/ CAN_A.CR.R = 0x04DB0006; /* Configure for 8MHz OSC, 100kHz bit time */ for (i=0; i<64; i++) { CAN_A.BUF[i].CS.B.CODE = 0; /* Inactivate all message buffers */ } CAN_A.BUF[0].CS.B.CODE = 8; /* Message Buffer 0 set to TX INACTIVE */ /* Use 1 pair of the next four lines of code for MPC551x or MPC555x */ /*SIU.PCR[48].R = 0x062C;*/ /* MPC551x: Configure pad as CNTXA, open drain */ /*SIU.PCR[49].R = 0x0500;*/ /* MPC551x: Configure pad as CNRXA */ SIU.PCR[83].R = 0x062C; /* MPC555x: Configure pad as CNTXA, open drain */ SIU.PCR[84].R = 0x0500; /* MPC555x: Configure pad as CNRXA */ CAN_A.MCR.R = 0x0000003F; /* Negate FlexCAN A halt state for 64 MB */ } void initCAN_C (void) { uint8_t i; CAN_C.MCR.R = 0x5000003F; /* Put in Freeze Mode & enable all 64 msg buffers*/ CAN_C.CR.R = 0x04DB0006; /* Configure for 8MHz OSC, 100kHz bit time */ for (i=0; i<64; i++) { CAN_C.BUF[i].CS.B.CODE = 0; /* Inactivate all message buffers */ } CAN_C.BUF[4].CS.B.IDE = 0; /* MB 4 will look for a standard ID */ CAN_C.BUF[4].ID.B.STD_ID = 555; /* MB 4 will look for ID = 555 */ CAN_C.BUF[4].CS.B.CODE = 4; /* MB 4 set to RX EMPTY */ CAN_C.RXGMASK.R = 0x1FFFFFFF; /* Global acceptance mask */ /* Use 1 pair of the next four lines of code for MPC551x or MPC555x */ /*SIU.PCR[52].R = 0x062C;*/ /* MPC551x: Configure pad as CNTXC, open drain */ /*SIU.PCR[53].R = 0x0500;*/ /* MPC551x: Configure pad as CNRXC */ SIU.PCR[87].R = 0x0E2C; /* MPC555x: Configure pad as CNTXC, open drain */ SIU.PCR[88].R = 0x0D00; /* MPC555x: Configure pad as CNRXC */ CAN_C.MCR.R = 0x0000003F; /* Negate FlexCAN C halt state for 64 MB */ } Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 213 void TransmitMsg (void) { uint8_ti; } /* Assumption: Message buffer CODE is INACTIVE */ const uint8_t TxData[] = {"Hello"}; /* Transmit string*/ CAN_A.BUF[0].CS.B.IDE = 0; /* Use standard ID length */ CAN_A.BUF[0].ID.B.STD_ID = 555; /* Transmit ID is 555 */ CAN_A.BUF[0].CS.B.RTR = 0; /* Data frame, not remote Tx request frame */ CAN_A.BUF[0].CS.B.LENGTH = sizeof(TxData) -1 ; /* # bytes to transmit w/o null */ for (i=0; i<sizeof(TxData); i++) { CAN_A.BUF[0].DATA.B[i] = TxData[i]; /* Data to be transmitted */ } CAN_A.BUF[0].CS.B.SRR = 1; /* Tx frame (not req'd for standard frame)*/ CAN_A.BUF[0].CS.B.CODE =0xC; /* Activate msg. buf. to transmit data frame */ void receiveMsg (void) { uint8_t j; uint32_t dummy; /* Use 1 of the next 2 lines: /*while (CAN_C.IFLAG1.B.BUF04I == 0) {};*//* MPC551x: Wait for CAN C MB 4 flag */ while (CAN_C.IFFRL.B.BUF04I == 0) {}; *//* MPC555x: Wait for CAN C MB 4 flag */ RxCODE = CAN_C.BUF[4].CS.B.CODE; /* Read CODE, ID, LENGTH, DATA, TIMESTAMP*/ RxID = CAN_C.BUF[4].ID.B.STD_ID; RxLENGTH = CAN_C.BUF[4].CS.B.LENGTH; for (j=0; j<RxLENGTH; j++) { RxDATA[j] = CAN_C.BUF[4].DATA.B[j]; } RxTIMESTAMP = CAN_C.BUF[4].CS.B.TIMESTAMP; dummy = CAN_C.TIMER.R; /* Read TIMER to unlock message buffers */ /* Use 1 of the next 2 lines: */ /*CAN_C.IFLAG1.R = 0x00000010;*/ /* MPC551x: Clear CAN C MB 4 flag */ CAN_C.IFRL.R = 0x00000010; /* MPC555x: Clear CAN C MB 4 flag */ } void main(void) { volatile uint32_t IdleCtr = 0; } initCAN_C(); initCAN_A(); TransmitMsg(); receiveMsg(); while (1) { IdleCtr++; } /* /* /* /* /* Initialize FLEXCAN C Initialize FlexCAN A Transmit one message Wait for the message Idle loop: increment & one of its buffers for receive*/ & one of its buffers for transmit*/ from a FlexCAN A buffer */ to be received at FlexCAN C */ counter */ Qorivva Simple Cookbook, Rev. 4 214 Freescale Semiconductor 25.3.2 MPC56xxB/S (MPC56xxB shown) /* main.c: FlexCAN program */ /* Description: Transmit one message from FlexCAN 0 buf. 0 to FlexCAN C buf. 1 */ /* Rev 0.1 Jan 16, 2006 S.Mihalik, Copyright Freescale, 2006. All Rights Reserved */ /* Rev 0.2 Jun 6 2006 SM - changed Flexcan A to C & enabled 64 msg buffers */ /* Rev 0.3 Jun 15 2006 SM - 1. Made globals uninitialized */ /* 2. RecieveMsg function: read CANx_TIMER, removed setting buffer's CODE*/ /* 3. added idle loop code for smoother Nexus trace */ /* 4. modified for newer Freescale header files (r 16) */ /* Rev 0.4 Aug 11 2006 SM - Removed redundant CAN_A.MCR init */ /* Rev 0.5 Jan 31 2007 SM - Removed other redundant CAN_C.MCR init */ /* Rev 0.6 Mar 08 2007 SM - Corrected init of MBs- cleared 64 MBs, instead of 63 */ /* Rev 0.7 Jul 20 2007 SM - Changes for MPC5510 */ /* Rev 0.8 May 15 2008 SM - Changes for new header file symbols */ /* Rev 0.9 May 22 2009 SM - Changes for MPC56xxB/P/S */ /* Rev 1.0 Jul 10 2009 SM - Simplified, cleared CAN Msg Buf flag by writing to reg */ /* not bit, increased Tx pads slew rate, chg’d RxCODE, RxLENGTH, dummy data types*/ /* & init receiving CAN first to allow CAN bus sync time before receiving 1st msg*/ /* Rev 1.1 Mar 14 2010 SM - modified initModesAndClock, updated header file */ /* NOTE!! structure canbuf_t in jdp.h header file modified to allow byte addressing*/ #include "MPC5604B_0M27V_0102.h" /* Use proper include file */ uint32_t RxCODE; uint32_t RxID; uint32_t RxLENGTH; uint8_t RxDATA[8]; uint32_t RxTIMESTAMP; /* /* /* /* /* Received Received Recieved Received Received message message message message message buffer code */ ID */ number of data bytes */ data string*/ time */ void initModesAndClks(void) { ME.MER.R = 0x0000001D; /* Enable DRUN, RUN0, SAFE, RESET modes */ /* Initialize PLL before turning it on: */ /* Use 1 of the next 2 lines depending on crystal frequency: */ CGM.FMPLL_CR.R = 0x02400100; /* 8 MHz xtal: Set PLL0 to 64 MHz */ /*CGM.FMPLL_CR.R = 0x12400100;*/ /* 40 MHz xtal: Set PLL0 to 64 MHz */ ME.RUN[0].R = 0x001F0074; /* RUN0 cfg: 16MHzIRCON,OSC0ON,PLL0ON,syclk=PLL */ ME.RUNPC[1].R = 0x00000010; /* Peri. Cfg. 1 settings: only run in RUN0 mode */ ME.PCTL[16].R = 0x01; /* MPC56xxB/P/S FlexCAN0: select ME.RUNPC[1] */ ME.PCTL[17].R = 0x01; /* MPC56xxB/S FlexCAN1: select ME.RUNPC[1] */ ME.PCTL[68].R = 0x01; /* MPC56xxB/S SIUL: select ME.RUNPC[1] */ /* Mode Transition to enter RUN0 mode: */ /* Enter RUN0 Mode & Key */ /* Enter RUN0 Mode & Inverted Key */ /* Wait for mode transition to complete */ /* Note: could wait here using timer and/or I_TC IRQ */ while(ME.GS.B.S_CURRENTMODE != 4) {} /* Verify RUN0 is the current mode */ ME.MCTL.R = 0x40005AF0; ME.MCTL.R = 0x4000A50F; while (ME.GS.B.S_MTRANS) {} } void initPeriClkGen(void) { CGM.SC_DC[1].R = 0x80; } /* MPC56xxB/S: Enable peri set 2 sysclk divided by 1 */ void disableWatchdog(void) { SWT.SR.R = 0x0000c520; /* Write keys to clear soft lock bit */ SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; /* Clear watchdog enable (WEN) */ } void initCAN_1 (void) { uint8_t i; CAN_1.MCR.R = 0x5000003F; CAN_1.CR.R = 0x04DB0006; for (i=0; i<64; i++) { CAN_1.BUF[i].CS.B.CODE = 0; } CAN_1.BUF[4].CS.B.IDE = 0; CAN_1.BUF[4].ID.B.STD_ID = 555; CAN_1.BUF[4].CS.B.CODE = 4; /* Put in Freeze Mode & enable all 64 msg bufs */ /* Configure for 8MHz OSC, 100kHz bit time */ /* Inactivate all message buffers */ /* MB 4 will look for a standard ID */ /* MB 4 will look for ID = 555 */ /* MB 4 set to RX EMPTY */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 215 } CAN_1.RXGMASK.R = 0x1FFFFFFF; SIU.PCR[42].R = 0x0624; SIU.PCR[35].R = 0x0100; SIU.PSMI[0].R = 0x00; CAN_1.MCR.R = 0x0000003F; /* Global acceptance mask */ /* MPC56xxB: Config port C10 as CAN1TX open drain */ /* MPC56xxB: Configure port C3 as CAN1RX */ /* MPC56xxB: Select PCR 35 for CAN1RX Input */ /* Negate FlexCAN 1 halt state for 64 MB */ void initCAN_0 (void) { uint8_t i; } CAN_0.MCR.R = 0x5000003F; CAN_0.CR.R = 0x04DB0006; for (i=0; i<64; i++) { CAN_0.BUF[i].CS.B.CODE = 0; } CAN_0.BUF[0].CS.B.CODE = 8; SIU.PCR[16].R = 0x0624; SIU.PCR[17].R = 0x0100; CAN_0.MCR.R = 0x0000003F; /* Put in Freeze Mode & enable all 64 msg bufs */ /* Configure for 8MHz OSC, 100kHz bit time */ /* Inactivate all message buffers */ /* Message Buffer 0 set to TX INACTIVE */ /* MPC56xxB: Config port B0 as CAN0TX, open drain */ /* MPC56xxB: Configure port B1 as CAN0RX */ /* Negate FlexCAN 0 halt state for 64 MB */ void TransmitMsg (void) { uint8_ti; } /* Assumption: Message buffer CODE is INACTIVE */ const uint8_t TxData[] = {"Hello"}; /* Transmit string*/ CAN_0.BUF[0].CS.B.IDE = 0; /* Use standard ID length */ CAN_0.BUF[0].ID.B.STD_ID = 555; /* Transmit ID is 555 */ CAN_0.BUF[0].CS.B.RTR = 0; /* Data frame, not remote Tx request frame */ CAN_0.BUF[0].CS.B.LENGTH = sizeof(TxData) -1 ; /* # bytes to transmit w/o null */ for (i=0; i<sizeof(TxData); i++) { CAN_0.BUF[0].DATA.B[i] = TxData[i]; /* Data to be transmitted */ } CAN_0.BUF[0].CS.B.SRR = 1; /* Tx frame (not req'd for standard frame)*/ CAN_0.BUF[0].CS.B.CODE =0xC; /* Activate msg. buf. to transmit data frame */ void RecieveMsg (void) { uint8_t j; uint32_t dummy; } while (CAN_1.IFRL.B.BUF04I == 0) {}; /* Wait for CAN 1 MB 4 flag */ RxCODE = CAN_1.BUF[4].CS.B.CODE; /* Read CODE, ID, LENGTH, DATA, TIMESTAMP */ RxID = CAN_1.BUF[4].ID.B.STD_ID; RxLENGTH = CAN_1.BUF[4].CS.B.LENGTH; for (j=0; j<RxLENGTH; j++) { RxDATA[j] = CAN_1.BUF[4].DATA.B[j]; } RxTIMESTAMP = CAN_1.BUF[4].CS.B.TIMESTAMP; dummy = CAN_1.TIMER.R; /* Read TIMER to unlock message buffers */ CAN_1.IFRL.R = 0x00000010; /* Clear CAN 1 MB 4 flag */ void main(void) { volatile uint32_t IdleCtr = 0; } initModesAndClks(); initPeriClkGen(); disableWatchdog(); initCAN_1(); initCAN_0(); TransmitMsg(); RecieveMsg(); while (1) { IdleCtr++; } /* Initialize mode entries */ /* Initialize peripheral clock generation for DSPIs */ /* Disable watchdog */ /* Initialize FLEXCAN 1 & one of its buffers for receive*/ /* Initialize FlexCAN 0 & one of its buffers for transmit*/ /* Transmit one message from a FlexCAN 0 buffer */ /* Wait for the message to be recieved at FlexCAN 1 */ /* Idle loop: increment counter */ Qorivva Simple Cookbook, Rev. 4 216 Freescale Semiconductor 25.3.3 MPC56xxP /* main.c: FlexCAN program */ /* Description: Transmit one message from FlexCAN 0 buf. 0 */ /* Rev 0.1 Jan 16, 2006 S.Mihalik, Copyright Freescale, 2006. All Rights Reserved */ /* Rev 0.2 Jun 6 2006 SM - changed Flexcan A to C & enabled 64 msg buffers */ /* Rev 0.3 Jun 15 2006 SM - 1. Made globals uninitialized */ /* 2. RecieveMsg function: read CANx_TIMER, removed setting buffer's CODE*/ /* 3. added idle loop code for smoother Nexus trace */ /* 4. modified for newer Freescale header files (r 16) */ /* Rev 0.4 Aug 11 2006 SM - Removed redundant CAN_A.MCR init */ /* Rev 0.5 Jan 31 2007 SM - Removed other redundant CAN_C.MCR init */ /* Rev 0.6 Mar 08 2007 SM - Corrected init of MBs- cleared 64 MBs, instead of 63 */ /* Rev 0.7 Jul 20 2007 SM - Changes for MPC5510 */ /* Rev 0.8 May 15 2008 SM - Changes for new header file symbols */ /* Rev 0.9 May 22 2009 SM - Changes for MPC56xxB/P/S */ /* Rev 1.0 Jul 10 2009 SM - Changes made for MPC56xxP, simplified. */ /* Cleared CAN Msg Buf flag by writing to reg. not bit, */ /* increased Tx pads slew rate, changed RxCODE, RxLENGTH, dummy data types and */ /* init receiving CAN first to allow CAN bus sync time before receiving first msg*/ /* Rev 1.0 Mar 14 1020 SM - Modified initModesAndClks, updated header */ /* NOTE!! structure canbuf_t DATA in header file modified to allow byte addressing*/ #include "Pictus_Header_v1_09.h" /* Use proper include file */ uint32_t RxCODE; uint32_t RxID; uint32_t RxLENGTH; uint8_t RxDATA[8]; uint32_t RxTIMESTAMP; /* /* /* /* /* Received Received Recieved Received Received message message message message message buffer code */ ID */ number of data bytes */ data string*/ time */ void initModesAndClks(void) { ME.MER.R = 0x0000001D; /* Enable DRUN, RUN0, SAFE, RESET modes */ /* Initialize PLL before turning it on: */ /* Use 2 of the next 4 lines depending on crystal frequency: */ /*CGM.CMU_0_CSR.R = 0x000000004;*/ /* Monitor FXOSC > FIRC/4 (4MHz); no PLL monitor*/ /*CGM.FMPLL[0].CR.R = 0x02400100;*/ /* 8 MHz xtal: Set PLL0 to 64 MHz */ CGM.CMU_0_CSR.R = 0x000000000; /* Monitor FXOSC > FIRC/1 (16MHz); no PLL monitor*/ CGM.FMPLL[0].CR.R = 0x12400100; /* 40 MHz xtal: Set PLL0 to 64 MHz */ ME.RUN[0].R = 0x001F0074; /* RUN0 cfg: 16MHzIRCON,OSC0ON,PLL0ON,syclk=PLL */ ME.RUNPC[1].R = 0x00000010; /* Peri. Cfg. 1 settings: only run in RUN0 mode*/ ME.PCTL[16].R = 0x01; /* MPC56xxB/P/S FlexCAN0: select ME.RUNPC[1] */ ME.PCTL[26].R = 0x01; /* MPC56xxP SafetyPort: select ME.RUNPC[1] */ /* Mode Transition to enter RUN0 mode: */ ME.MCTL.R = 0x40005AF0; /* Enter RUN0 Mode & Key */ ME.MCTL.R = 0x4000A50F; /* Enter RUN0 Mode & Inverted Key */ while (ME.GS.B.S_MTRANS) {} /* Wait for mode transition to complete */ /* Note: could wait here using timer and/or I_TC IRQ */ while(ME.GS.B.S_CURRENTMODE != 4) {} /* Verify RUN0 is the current mode */ } void initPeriClkGen(void) { CGM.AC2SC.R = 0x04000000; CGM.AC2DC.R = 0x80000000; } /* MPC56xxP Safety Port: Select PLL0 for aux clk 0 */ /* MPC56xxP Safety Port: Enable aux clk 0 div by 1 */ void disableWatchdog(void) { SWT.SR.R = 0x0000c520; /* Write keys to clear soft lock bit */ SWT.SR.R = 0x0000d928; SWT.CR.R = 0x8000010A; /* Clear watchdog enable (WEN) */ } void initSAFEPORT (void) { uint8_t i; SAFEPORT.MCR.R = 0x5000001F; /* Put in Freeze Mode & enable all 32 msg bufs */ /* Use 1 of the next 2 lines depending on crystal frequency: */ /*SAFEPORT.CR.R = 0x04DB0006; */ /* Configure for 8MHz OSC, 100kHz bit time */ SAFEPORT.CR.R = 0x18DB0006; /* Configure for 40MHz OSC, 100kHz bit time */ for (i=0; i<32; i++) { SAFEPORT.BUF[i].CS.B.CODE = 0; /* Inactivate all message buffers */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 217 } } SAFEPORT.BUF[4].CS.B.IDE = 0; /* MB 4 will look for a standard ID */ SAFEPORT.BUF[4].ID.B.STD_ID = 555; /* MB 4 will look for ID = 555 */ SAFEPORT.BUF[4].CS.B.CODE = 4; /* MB 4 set to RX EMPTY */ SAFEPORT.RXGMASK.R = 0x1FFFFFFF; /* Global acceptance mask */ SIU.PCR[14].R = 0x0624; /* MPC56xxP: Config port A14 as CAN1TX, open drain */ SIU.PCR[15].R = 0x0900; /* MPC56xxP: Configure port A15 as CAN1RX */ SAFEPORT.MCR.R = 0x0000001F; /* Negate SAFETY PORT halt state for 32 MB */ void initCAN_0 (void) { uint8_t i; CAN_0.MCR.R = 0x5000001F; /* Put in Freeze Mode & enable all 32 msg bufs */ /* Use 1 of the next 2 lines depending on crystal frequency: */ /*CAN_0.CR.R = 0x04DB0006; */ /* Configure for 8MHz OSC, 100kHz bit time */ CAN_0.CR.R = 0x18DB0006; /* Configure for 40MHz OSC, 100kHz bit time */ for (i=0; i<32; i++) { CAN_0.BUF[i].CS.B.CODE = 0; /* Inactivate all message buffers */ } CAN_0.BUF[0].CS.B.CODE = 8; /* Message Buffer 0 set to TX INACTIVE */ SIU.PCR[16].R = 0x0624; /* MPC56xxP: Config port B0 as CAN0TX, open drain */ SIU.PCR[17].R = 0x0500; /* MPC56xxP: Configure port B1 as CAN0RX */ CAN_0.MCR.R = 0x0000001F; /* Negate FlexCAN 0 halt state for 21 MB */ } void TransmitMsg (void) { uint8_ti; } /* Assumption: Message buffer CODE is INACTIVE */ const uint8_t TxData[] = {"Hello"}; /* Transmit string*/ CAN_0.BUF[0].CS.B.IDE = 0; /* Use standard ID length */ CAN_0.BUF[0].ID.B.STD_ID = 555; /* Transmit ID is 555 */ CAN_0.BUF[0].CS.B.RTR = 0; /* Data frame, not remote Tx request frame */ CAN_0.BUF[0].CS.B.LENGTH = sizeof(TxData) -1 ; /* # bytes to transmit w/o null */ for (i=0; i<sizeof(TxData); i++) { CAN_0.BUF[0].DATA.B[i] = TxData[i]; /* Data to be transmitted */ } CAN_0.BUF[0].CS.B.SRR = 1; /* Tx frame (not req'd for standard frame)*/ CAN_0.BUF[0].CS.B.CODE =0xC; /* Activate msg. buf. to transmit data frame */ void RecieveMsg (void) { uint8_t j; uint32_t dummy; } while (SAFEPORT.IFRL.B.BUF04I == 0) {}; /* Wait for SAFETY PORT MB 4 flag */ RxCODE = SAFEPORT.BUF[4].CS.B.CODE; /* Read CODE, ID, LENGTH, DATA, TIMESTAMP */ RxID = SAFEPORT.BUF[4].ID.B.STD_ID; RxLENGTH = SAFEPORT.BUF[4].CS.B.LENGTH; for (j=0; j<RxLENGTH; j++) { RxDATA[j] = SAFEPORT.BUF[4].DATA.B[j]; } RxTIMESTAMP = SAFEPORT.BUF[4].CS.B.TIMESTAMP; dummy = SAFEPORT.TIMER.R; /* Read TIMER to unlock message buffers */ SAFEPORT.IFRL.R = 0x00000010; /* Clear SAFETY PORT MB 4 flag */ void main(void) { volatile uint32_t IdleCtr = 0; initModesAndClks(); initPeriClkGen(); disableWatchdog(); initSAFEPORT(); initCAN_0(); TransmitMsg(); RecieveMsg(); while (1) { IdleCtr++; } /* Initialize mode entries */ /* Initialize peripheral clock generation for DSPIs */ /* Disable watchdog */ /* Initialize SafetyPort & one of its buffers for receive*/ /* Initialize FlexCAN 0 & one of its buffers for transmit*/ /* Transmit one message from a FlexCAN 0 buffer */ /* Wait for the message to be recieved at FlexCAN 1 */ /* Idle loop: increment counter */ Qorivva Simple Cookbook, Rev. 4 218 Freescale Semiconductor 26 Flash: Configuration 26.1 Description Task: Configure the flash performance parameters for a 64 MHz system clock and increase performance for branch instructions by enabling branch target buffers and branch prediction. Key elements of the flash module include the flash array and its line buffers. Line buffers allow overlapping fast access between the crossbar and a flash module line buffer with a slower access between a line buffer and the flash array. Line buffer bus interface width is 32 or 64 bits, but the actual buffer width is the width of the flash array: either 128 or 256 bits. Hence one line buffer fill from the array will provide multiple bus interface transfers to, for example, the core. Parameters affecting transfers between the flash array and the line buffers, such as read wait states, are not optimal out of reset and for best performance should be configured for desired target a system clock frequency. When modifying characteristics for a memory, such as in this example, it is good practice not to execute code in the same memory that is having its characteristics modified. Hence code to modify the flash performance parameters of the module’s configuration register will be executed from SRAM. For a more complete checklist of performance items, see application note AN3519, “Optimizing Performance for the MPC5500 Family.” MPC5500 e200 Core other crossbar bus masters Branch Target Buffers Crossbar Flash Module Configuration Reg. 32 or 64 bits Line Buffer Line Buffer SRAM other crossbar bus slaves ... other line buffers 128 or 256 bits Flash Array Figure 39. Flash Configuration Example Exercise: Change configuration parameters and attempt to determine performance differences. Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 219 26.2 Design Flash configuration consists of optimizing the flash bus interface for the following areas: • Enabling line buffers • Number of wait states and pipeline hold cycles for a given frequency • Line buffer prefetch controls • Line buffer configurations (MPC551x and MPC56xxB/P/S only) • Read-While-Write control (MPC56xxB/P/S only) NOTE: Be sure to verify flash configuration parameters with the latest data sheet for each device. 26.2.1 Line Buffer Enable BFEN or BFE (FBIU Line Read Buffers Enable): Enables or disables line read buffer hits. Recommendation: Set BFEN = 1 to enable the use of line buffers. 26.2.2 Line Buffer Wait State and Hold Cycles Wait states and hold cycles are defined to be between the flash line buffers and flash array, not between flash and core or crossbar. RWSC (Read Wait State Control): Defines the number of wait states to be added to the flash array access time for reads. WWSC (Write Wait State Control): Defines the number of wait states to be added to the flash array access time for writes. APC (Address Pipelining Control): Address pipelining drives a subsequent access address and control signals while waiting for the current access to complete. APC defines the number of hold cycles between flash array access requests. Recommendation: Use the parameter values in the reference manuals for the desired system clock frequency to get fastest performance. 26.2.3 Line Buffer Prefetch Controls When one line buffer is being accessed from the crossbar, another line buffer can be prefetching the next sequential address from the flash array at the same time. This overlapping operation allows the flash module to provide sequential address accesses without any delay. Prefetching does not provide any benefit if accesses are not sequential beyond a flash line. Therefore as a general rule, prefetching should be enabled when accesses are expected to be sequential, such as instructions. Generally data access has addresses more random than sequential, so prefetching will not be a benefit. IPFEN or IPFE (Instruction Prefetch Enable): Enables line buffers to prefetch instructions. Qorivva Simple Cookbook, Rev. 4 220 Freescale Semiconductor Recommendation: Since most instruction accesses are sequential, instruction prefetching should be enabled for any access. DPFEN or DPFE (Data Prefetch Enable): Enables line buffers to prefetch data. Recommendation: Determining the setting is not as obvious as for instruction prefetch enable. Many data accesses may be sequential, such as table data, filter coefficients, and for MPC5606S, large graphic objects. For engine control, start with enabled for any read access. For MPC56xxB/P, start with DPFEN = 0 and for MPC5606S with graphics in internal flash, start with DPFEN = 1. PFLIM (Prefetch Limit): Defines the maximum number of prefetches done on a buffer miss. Recommendation: Prefetch the next sequential line on a hit or a miss. 26.2.4 Master Prefetch Enables / Disables MnPFE (Master n Pre Fetch Enable) and MnPFD (MPC56xxB/P/S — Disable): Enables (disables on MPC56xxB/P/S) line buffers to prefetch an entire line in the flash array to a line buffer for that master. The following table summarizes values for given devices from the respective device documentation. Recommendation: Most of the time, only enable prefetching for core instructions. This is based on the rationale that, most of the time, prefetching the next sequential address usually benefits only instruction fetching. Few prefetch benefits occur from other masters, such as data fetching, external bus master, and DMA. Most of the time DMA is simply moving a byte, half word, or word for MPC5500 and MPC5600 devices. (However, this may not be true for bus masters such as a display controller, FlexRAY, or FEC. Settings may be optimized differently in these cases.) Recommend only prefetching for master 0 (core). 26.2.5 Line Buffer Configuration (MPC551x and MPC56xxB/P/S) Some devices have two interfaces (ports) to the flash module. Each port has its own set of line buffers; hence there are two configuration registers to initialize. Line buffers for each port can be organized as a “pool” available for data and instructions, or with fixed partitions. LBCFG or BCFG (Line Buffer Configuration): Specifies configuration of how many buffers are used for instruction fetches and data fetches. Recommendation: Generally the optimal configuration s would be to have a fixed partition of mostly instructions fetch line buffers, or a pool for both instruction and data fetches. On the MPC551x, the line buffer control value is loaded initially from the shadow block by BAM code. 26.2.6 Read-While-Write (RWW) Control (MPC56xxB/P/S) Low cost flash arrays such as on MPC56xxB/P/S do not support RWW as in MPC55xx devices. RWWC (Read-While-Write Control): Defines the flash controller response to flash reads while the array is busy with a program (write) erase operation. Likely Recommendation: Terminate RWW or generate bus stall with abort and abort interrupt enabled so the CPU knows it occurred. Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 221 Table 102. MPC56xxB/P/S Flash Configuration Example Values for 64 MHz System Clock General Recommendations Access Parameter Code Flash (MPC56xxB/P: Bank 0, MPC56xxS: Banks 0 & 2) 4 Line Buffers Per Port Parameter Symbol in register PFCR0 Result value: 0x1084_126F Port 0 Page Buffer Enable (Connected to Instruction Prefetch all masters on Enable MPC56xxB/P, connected to core only on MPC56xxS) Data Prefetch Enable Comments Data Flash (Bank 1) 1 Line Buffer Per Port Parameter Symbol in register PFCR1 Result Value: 0x1084_0101 Comments B0_P0_BFE = 1 Enable port’s buffers B1_P0_BFE = 1 Enable port’s buffer B0_P0_IPFE = 1 Instructions are mostly sequential, so prefetching can improve performance. - B0_P0_DPFE = 0 Data accesses are expected to generally be random, not sequential - Prefetch Limit B0_P0_PFLIM = 3 Prefetch on hit or miss - - Page Buffer Configuration B0_P0_BCFG = 3 Allocate 3 line buffers for instructions, 1 for data - B0_P1_BFE = 1 Enable port’s buffers B1_P1_BFE = 1 Enable port’s buffer B0_P1_IPFE = 0 No instruction access on port 1 - B0_P1_DPFE = 1 Enable prefetching assuming there is significant sequential data - Prefetch Limit B0_P1_PFLIM = 1 Prefetch on miss only (allows more bandwidth for core) - Page Buffer Configuration B0_P1_BCFG = 0 All 4 line buffers available for any access - - BK0_RWSC = 2 Values are system clock frequency dependent BK1_RWSC = 2 Values are system clock frequency dependent Port 1 Page Buffer Enable (Connected to Instruction Prefetch DCU, DMA on Enable MPC56xxS) Data Prefetch Port 1 fields Enable ignored in MPC56xxB/P) Array Access Read Wait States (for 64 MHz) Write Wait States BK0_WWSC = 2 Adv. Pipeline Ctl. BK0_APC = 2 Read While Write Ctl. BK0_RRWC = 0 BK1_WWSC = 2 BK1_APC = 2 Terminate RWW BK1_RRWC = 0 attempt with error response. Assumes software must first check if any program or erase commands are in progress. Terminate RWW attempt with error response. Assumes software must first check if any program or erase commands are in progress. Qorivva Simple Cookbook, Rev. 4 222 Freescale Semiconductor Table 103. MPC56xxB/P/S Crossbar Master Assignments. Crossbar Physical Master ID MPC56xxB MPC56xxP MPC56xxS 0 e200z0h core instructions e200z0h core instructions e200z0h core instructions 1 e200z0h core data e200z0h core data e200z0h core data 2 - eDMA eDMA 3 - FlexRAY DCU 4 to 7 - - - Table 104. MPC56xxB/P/S Access and Protection Example Settings (Result value for MPC56xxS in PFAPR = 0x03F2 005D) Parameter Symbol MPC56xxB Value in Register PFAPR Result Value: MPC56xxP Value in Register PFAPR Result Value: MPC56xxS Value in Register PFAPR Result Value: 0x00FE 000D 0x00FE 005D 0x03F2 005D Arbitration Mode ARBM 3 (start with round-robin and change after application testing as needed) Master n Prefetch Disable M0PFD 0 (since core instructions are mostly sequential, so prefetching should help) M1PFD 1 (assuming CPU data is not normally accessed sequentially) Master n Access Protection M2PFD - 1 (assuming eDMA will not 0 assuming eDMA will have have significant consecutive significant consecutive data accesses) used for graphics M3PFD - 1 (assuming FlexRAY will not have significant consecutive accesses) 0 (assuming DCU will have significant consecutive data used for graphics) M0AP 1 (R only for core instructions) M1AP 3 (R+W access allowed for core data) M2AP - M3AP - 1 (R only for eDMA) 1 (R only unless flashing over FlexRay) 1 (R only for DCU) Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 223 x Table 105. MPC551x, MPC55xx Flash Configuration Example Values (MPC551x port 0 configuration register also uses values LBCF=0, ARB=1 and PRI=0 here and MPC5634M configuration register also uses GCE = 0.) MnPFE Resulting Core Configuration APC RWSC WWSC DPFEN IPFEN PFLIM BFEN (others Register and 0x0) Value Device & Reference for APC, RWSC, WWSC Max. Target Frequency MPC5510 - port 0 (Reference: MPC5510 Data Sheet, Rev. 3 & MPC5510 Ref Manual, Rev. 1) Reset Default 0 7 7 3 0 0 0 0 PFCRP0 = 0x0800 FF00 Up to 25 MHz 1 0 0 1 0 1 2 1 PFCRP0 = 0x0801 0815 Up to 50 MHz 1 1 1 1 0 1 2 1 PFCRP0 = 0x0801 2915 Up to 80 MHz 1 2 2 1 0 1 2 1 PFCRP0 = 0x0801 4A15 MPC5533, MPC5534 Reset Default (Reference: MPC5534 Rev. 1 Addendum, Rev. 1 Up to 25 MHz 0 7 7 3 0 0 0 0 BUICR = 0x0000 FF00 1 0 0 1 1 1 2 1 BUICR = 0x0001 0855 1 1 1 1 1 1 2 1 BUICR = 0x0001 2955 1 2 2 1 1 1 2 1 BUICR = 0x0001 4A55 Up to 80 MHz 1 3 3 1 1 1 2 1 BUICR = 0x0001 6B55 Reset Default 0 7 7 3 0 0 0 0 BUICR = 0x0000 FF00 Up to 82 MHz 1 1 1 1 3 3 2 1 BUICR = 0x0001 29FD Up to 102 MHz 1 1 2 1 3 3 2 1 BUICR = 0x0001 2AFD Up to 132 MHz (MPC555x) or up to 135 MHz (MPC556x) 1 2 3 1 3 3 2 1 BUICR = 0x0001 4BFD Up to 147 MHz (MPC5566 only) 1 3 4 1 3 3 2 1 BUICR = 0x0001 6CFD MPC5632M, MPC5634M, Reset Default MPC5634M (Reference: MPC5634M Ref. Up to 40 MHz Manual, Rev. 2) 0 7 7 3 0 0 0 0 PFCR1 = 0x0000 FF00 1 1 1 1 1 1 2 1 PFCR1 = 0x0001 2955 Up to 62 MHz 1 2 2 1 1 1 2 1 PFCR1 = 0x0001 4A55 Up to 82 MHz 1 3 3 1 1 1 2 1 PFCR1 = 0x0001 6B55 (Note on up to 25 MHz row: This APC/RWSC/WWSC Up to 50 MHz combination requires setting the Flash MCR register bit Up to 75 MHz PRD=1.) MPC5553, MPC5554, MPC5565, MPC5566, MPC5567 (Reference: published data sheets as of April 2009) Qorivva Simple Cookbook, Rev. 4 224 Freescale Semiconductor Table 106. Flash Configuration Steps (Shown for MPC5554 for up to 82 MHz sysclk) Relevant Bit Fields Step Pseudo Code Data Init Determine desired configuration data. In this case, assume an MPC555x running up to 82 MHz. FLASH_CONFIG_DATA = 0x0001 29FD FLASH_CONFIG_REG = FLASH_BIUCR Configure Flash In RAM, create array structure of machine code to write 32 bits data in r3 to address in r4, ensure the instruction completes, then return. Instruction code is: — stw r3, 0 (r4) — mbar — blr uint32_t mem_write_code[] = { 0x90640000, 0x7C0006AC, 0x4E800020 } Create a new typedef for function pointer that does not return anything (void) and passes two integers (in r4, then r5 per EABI). typedef void (*mem_write_code_ptr_t) (int, int) Call memory write code function: Cast memory_write_code as a function pointerq — then de-reference it, which converts it to a function which passes, in order: — Value to be written (goes into r3 per EABI) — — Memory address used for write (goes into r4 per EABI) (*((mem_write_code_ptr_t) mem_write_code)) (FLASH_CONFIG_DATA, &FLASH_CONFIG_ADDR) Enable Branch Acceleration and Branch Target Buffers Enable branch acceleration for forward and backward branches (reset default = enabled) HID0[PBRED]=0 spr HID0 = 0x0 (reset default) Invalidate branch target buffers and Enable branch target buffers BUCSR[BBFI]=1 BUCSR[BPEN]=1 spr BUCSR = 0x0000 0201 Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 225 26.3 /* /* /* /* /* /* Code (Shown for MPC5554 with 64 MHz sysclk) main.c - Example of Copyright Freescale Rev 0.1 Jun 18 2008 Rev 0.2 Oct 30 2008 Rev 0.3 May 20 2009 Rev 0.4 Sep 11 2009 flash configuration plus using branch target buffers */ Semiconductor, Inc 2008 All rights reserved. */ Dan and Steve Mihalik */ S. Mihalik - added mbar to mem_write_code */ S. Mihalik - updated Config. value */ S. Mihalik - corrected BUCSR value to 0x0201 */ #include "mpc5554.h" /* Include appropriate header file like mpc5554.h */ #define FLASH_CONFIG_DATA 0x001029FD #define FLASH_CONFIG_REG FLASH.BIUCR.R /* MPC5554 config. value for up to 82 MHz */ /* Flash config. register address */ asm void enable_accel_BTB(void) { li r0, 0 /* Enable branch acceleration (HID[PBRED]=0) */ mtHID0 r0 li r0, 0x0201 /* Invalidate Branch Target Buffers and enable them */ mtBUCSR r0 } int main(void) { uint32_t i=0; /* Dummy idle counter */ uint32_t mem_write_code [] = { 0x90640000, /* stw r3,(0)r4 machine code; writes r3 contents to addr in r4 */ 0x7C0006AC, /* mbar machine code: ensure prior store completed */ 0x4E800020 /* blr machine code: branches to return address in link register */ }; typedef void (*mem_write_code_ptr_t)(uint32_t, uint32_t); /* create a new type def: a func pointer called mem_write_code_ptr_t */ /* which does not return a value (void) */ /* and will pass two 32 bit unsigned integer values */ /* (per EABI, the first parameter will be in r3, the second r4) */ (*((mem_write_code_ptr_t)mem_write_code)) /* cast mem_write_code as func ptr*/ /* * de-references func ptr, i.e. converts to func*/ (FLASH_CONFIG_DATA, /* which passes integer (in r3) */ (uint32_t)&FLASH_CONFIG_REG); /* and address to write integer (in r4) */ } enable_accel_BTB(); /* Enable branch accel., branch target buffers */ while(1) {i++;} /* Wait forever */ Qorivva Simple Cookbook, Rev. 4 226 Freescale Semiconductor 26.4 Code (Shown for MPC56xxS with 64 MHz sysclk) /* main.c - Example of flash configuration plus using branch target buffers */ /* Copyright Freescale Semiconductor, Inc 2009 All rights reserved. */ /* Rev 0.1 May 22 2009 S. Mihalik - Initial version based on AN2865 example */ /* Rev 0.2 Sep 11 2009 S. Mihalik - corrected BUCSR value to 0x0201 */ /* Rev 0.3 Mar 11 2010 S. Mihalik - corrected FLASH_CONFIG_DATA to 0x1084126F */ #include "jdp.h" /* Include appropriate header file like mpc5554.h */ #define #define #define #define FLASH_CONFIG_DATA 0x1084126F /* MPC56xxS flash config value for 64 MHz */ FLASH_CONFIG_REG CFLASH.PFCR0.R /* Flash config. register address */ FLASH_ACCESS_PROT_DATA 0x03F2005D /* MPC56xxS flash access prot. value */ FLASH_ACCESS_PROT_REG CFLASH.FAPR.R /* Flash Access Prot. Reg. address */ asm void enable_accel_BTB(void) { li r0, 0 /* Enable branch acceleration (HID[PBRED]=0) */ mtHID0 r0 li r0, 0x0201 /* Invalidate Branch Target Buffers and enable them */ mtBUCSR r0 } int main(void) { uint32_t i=0; /* Dummy idle counter */ /* NOTE: Structures are default aligned on a boundary which is a multiple of */ /* the largest sized element, 4 bytes in this case. The first two */ /* instructions are 4 bytes, so the last instruction is duplicated to */ /* avoid the compiler adding padding of 2 bytes before the instruction. */ uint32_t mem_write_code_vle [] = { 0x54640000, /* e_stw r3,(0)r4 machine code: writes r3 contents to addr in r4 */ 0x7C0006AC, /* mbar machine code: ensure prior store completed */ 0x00040004 /* 2 se_blr's machine code: branches to return address in link reg.*/ }; typedef void (*mem_write_code_ptr_t)(uint32_t, uint32_t); /* create a new type def: a func pointer called mem_write_code_ptr_t */ /* which does not return a value (void) */ /* and will pass two 32 bit unsigned integer values */ /* (per EABI, the first parameter will be in r3, the second r4) */ (*((mem_write_code_ptr_t)mem_write_code_vle)) /* cast mem_write_code as func ptr*/ /* * de-references func ptr, i.e. converts to func*/ (FLASH_CONFIG_DATA, /* which passes integer (in r3) */ (uint32_t)&FLASH_CONFIG_REG); /* and address to write integer (in r4) */ (*((mem_write_code_ptr_t)mem_write_code_vle)) /* cast mem_write_code as func ptr*/ /* * de-references func ptr, i.e. converts to func*/ (FLASH_ACCESS_PROT_DATA, /* which passes integer (in r3) */ (uint32_t)&FLASH_ACCESS_PROT_REG); /* and address to write integer (in r4) */ } enable_accel_BTB(); /* Enable branch accel., branch target buffers */ while(1) {i++;} /* Wait forever */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 227 Appendix A Interrupt Alignment Summary The table below shows the alignment requirements for interrupts, interrupt tables, and interrupt handlers. Table names used in examples are listed for reference. Table 107. Address Locations and Alignment for Interrupt Tables, ISR’s, and Handlers Interrupt Type Core Interrupts Memory Section in Examples ivor_branch_table (MPC551x, MPC56xxB/P/S only) ivor_handlers MPC551x, MPC55xxB/P/S MPC555x Address Alignment Address Alignment IVPR0:19 4 KB (0x1000) N.A. N.A. N.A. N.A. IVPR0:15 + IVORx16:28 IVPR: 64 KB (0x1 0000) Handlers: 16 Bytes (0x10) 2 KB (0x800) INTC_IACKR0:20 [VTBA] 2 KB (0x800) 2 KB (0x800) above a 4 KB boundary IVPR0:15 64 KB (0x1 0000) INTC SW mode Interrupts Software Vector Mode: INTC_IACKR0:20 intc_sw_isr_vector_table [VTBA] INTC HW mode Interrupts Hardware Vector Mode: intc_hw_branch_table IVPR0:19 + 0x800 Qorivva Simple Cookbook, Rev. 4 228 Freescale Semiconductor Appendix B Single Core Build Files This section describes the build files used in this application note’s examples that use only a single core. NOTE: These examples assume 32 KB of internal SRAM. Some devices, such as MPC5604S, only have 24 KB internal SRAM, so adjustments are needed for link files in that case. B.1 Memory Map for Executing from Internal RAM To execute from SRAM, the memory map in Figure 40 is used. Only 32 KB RAM is assumed, so the program can run on the MPC5500 family member with the least amount of memory. From this build, there is no special boot code — the debugger scripts must initialize internal RAM and MMU, as well as setting the instruction pointer to the beginning of code. This memory map is intended as a general example for a single-core processor for both MPC551x and MPC555x devices. Therefore some memory section names are defined that do not exist for some builds. For example, the Decrementer example does not use the section “intc_hw_branch_table.” Similarly, section “ivor_branch_table” is only used for interrupts on MPC551x, and does not exist for MPC555x applications. 0x4000 0000 0x4000 3000 Memory Segment Names Memory Section Names interrupts_ram (12 KB) ivor_branch_table intc_hw_branch_table ivor_handlers internal_ram (19 KB) intc_sw_isr_vector_table MPC551x core interrupts (code, etc.) (data, etc.) 0x4000 7CFF stack_ram (1 KB) (stack) For INTC HW vector mode. If using this mode, table starts at: 0x4000 0800 (MPC551x) 0x4000 0000 (MPC555x) For INTC SW vector mode. If using this mode, table starts at 0x4000 3000. 0x4000 7FFF Figure 40. Memory Map for Executing from Internal SRAM (IVPR Value is Passed as 0x4000 0000) Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 229 B.1.1 CodeWarrior Link File for Execution from Internal RAM /* 5500_ram.lcf - Simple minimal MPC5500 link file using 32 KB SRAM */ /* Aug 30 2007 initial version */ /* May 09 208 SM: Put stack in it's own 1KB (0x400) memory segment */ MEMORY { interrupts_ram: org = 0x40000000, len = 0x3000 internal_ram: org = 0x40003000, len = 0x4C00 stack_ram: org = 0x40007C00, len = 0x0400 } SECTIONS { GROUP : { .ivor_branch_table : {} /* For MPC5516 core interrupts */ .intc_hw_branch_table ALIGN (2048) : {} /* For INTC in HW Vector Mode */ .ivor_handlers : {} /* Handlers for core interrupts */ } > interrupts_ram GROUP : { .intc_sw_isr_vector_table : {} /* For INTC in SW Vector Mode */ .text : { *(.text) *(.rodata) *(.ctors) *(.dtors) *(.init) *(.fini) *(.eini) . = (.+15); } .sdata2 : {} extab : {} extabindex : {} } > internal_ram GROUP : { .data (DATA) .sdata (DATA) .sbss (BSS) .bss (BSS) } > internal_ram : : : : {} {} {} {} } /* Freescale CodeWarrior compiler address designations */ _stack_addr = ADDR(stack_ram)+SIZEOF(stack_ram); _stack_end = ADDR(stack_ram); /* These are not currently being used _heap_addr = ADDR(.bss)+SIZEOF(.bss); _heap_end = ADDR(internal_ram)+SIZEOF(internal_ram); */ __IVPR_VALUE = ADDR(interrupts_ram); /* L2 SRAM Location (used for L2 SRAM initialization) */ L2SRAM_LOCATION = 0x40000000; Qorivva Simple Cookbook, Rev. 4 230 Freescale Semiconductor B.1.2 /* /* /* /* Diab Link File for Execution from Internal RAM 5500_ram.lin - Simple minimal MPC5500 link file for 32 KB SRAM */ Jul 05 2007 S.M. Initial version. */ May 09 2008 S.M. Put stack in separate memory segment Notes: 1. assumption: debug scripts will init SRAM, MMU, start vector */ MEMORY { /************************************************************************/ /* Address Length Use */ /************************************************************************/ /* 0x4000_0000 - 0x4000_2FFF 12 KB RAM - Interrupt area */ /* 0x4000_3000 - 0x4000_7BFF 15 KB RAM - Code and data */ /* 0x4000_7C00 - 0x4000_7FFF 1 KB RAM - Stack */ /************************************************************************/ interrupts_ram:org = 0x40000000, len = 0x3000 internal_ram:org = 0x40003000, len = 0x4C00 stack_ram: org = 0x40007C00, len = 0x0400 } SECTIONS { GROUP : { .ivor_branch_table : {} /* For MPC5516 core interrupts */ .intc_hw_branch_table ALIGN (2048): {} /* For INTC in HW Vector Mode */ .ivor_handlers : {} /* Handlers for core interrupts */ } > interrupts_ram GROUP : { .intc_sw_isr_vector_table ALIGN (2048): {} /* For INTC SW Vector Mode */ .text (TEXT) : { *(.text) *(.rodata) *(.ctors) *(.dtors) *(.init) *(.fini) *(.eini) . = (.+15) & ~15; } .sdata2 (TEXT) : {} } > internal_ram GROUP : { .data (DATA) LOAD(ADDR(.sdata2)+SIZEOF(.sdata2)) : {} .sdata (DATA) LOAD(ADDR(.sdata2)+SIZEOF(.sdata2)+SIZEOF(.data)) : {} .sbss (BSS) : {} .bss (BSS) : {} } > internal_ram } __IVPR_VALUE = ADDR(interrupts_ram); /* Pass address to be loaded to spr IVPR */ Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 231 __SP_INIT = ADDR(stack_ram)+SIZEOF(stack_ram); __SP_END = ADDR(stack_ram); __DATA_ROM = ADDR(.sdata2)+SIZEOF(.sdata2); __DATA_RAM = ADDR(.data); __DATA_END = ADDR(.sdata)+SIZEOF(.sdata); __BSS_START = ADDR(.sbss); __BSS_END = ADDR(.bss)+SIZEOF(.bss); __HEAP_START= ADDR(.bss)+SIZEOF(.bss); __HEAP_END = ADDR(internal_ram)+SIZEOF(internal_ram); Qorivva Simple Cookbook, Rev. 4 232 Freescale Semiconductor B.1.3 Green Hills Link File for Execution from Internal RAM /* 5500_ram.ld - Example minimal MPC5500 link file- 512 KB flash, 32 KB SRAM */ /* May 09 2008 S.M., G.L. Initial version. */ /* Notes: 1. assumption: debug scripts will init SRAM, MMU, start vector */ MEMORY { /*******************************************************************/ /* Address Length Use */ /*******************************************************************/ /* 4000_0000-4000_2fff 12 KB RAM Interrupt area */ /* 4000_3000-4000_7BFF 19 KB RAM Code and data except stack */ /* 4000_7C00-4000_7FFF 1 KB RAM Stack */ /*******************************************************************/ interrupts_ram internal_ram stack_ram : : : ORIGIN = 0x40000000, LENGTH = 12K ORIGIN = . , LENGTH = 19K ORIGIN = . , LENGTH = 1K } CONSTANTS { stack_reserve = 1K heap_reserve = 1K } SECTIONS { // RAM SECTIONS .ivor_branch_table : > interrupts_ram .intc_hw_branch_table ALIGN(2048) : > . .ivor_handlers : > . .init .text .syscall .rodata .sdata2 .secinfo .fixaddr .fixtype : : : : : : : : > > > > > > > > /* For MPC5516 core interrupts */ /* For INTC in HW Vector Mode */ /* For core interrupt handlers */ internal_ram . . . . . . . .PPC.EMB.sdata0 ABS .PPC.EMB.sbss0 CLEAR ABS .sdabase ALIGN(8) .sdata .sbss .data .bss .heap ALIGN(16) PAD(heap_reserve) : : : : : : : : > > > > > > > > /* For GHS hostio support */ /* For GHS startup code */ . . . . . . . . Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 233 .stack ALIGN(16) PAD(stack_reserve) : > stack_ram /* SP init value addr stack_ram + pad */ __STACK_SIZE __SP_END = stack_reserve; = ADDR(.stack); __IVPR_VALUE = MEMADDR(interrupts_ram); /* Pass address to be loaded to IVPR */ // // These special symbols mark the bounds of RAM and ROM memory. // They are used by the MULTI debugger. // __ghs_ramstart = MEMADDR(interrupts_ram); __ghs_ramend = MEMENDADDR(stack_ram); __ghs_romstart = 0; __ghs_romend = 0; // // These special symbols mark the bounds of RAM and ROM images of boot code. // They are used by the GHS startup code (_start and __ghs_ind_crt0). // __ghs_rambootcodestart = MEMADDR(interrupts_ram); __ghs_rambootcodeend = ADDR(.fixtype); __ghs_rombootcodestart = 0; __ghs_rombootcodeend = 0; } Qorivva Simple Cookbook, Rev. 4 234 Freescale Semiconductor B.2 Memory Map for Executing from Internal Flash To execute from flash, the memory map in Figure 41 is used. Only 512 KB flash is assumed, so the program can run on the MPC5500 family member with the least amount of memory. When executing from flash, we must add some “boot” code, including a Reset Configuration Half Word (RCHW), a starting vector, and code to initialize SRAM. This is put in the “boot” memory section. After reset, the BAM code inside the device will initialize the MMU. This memory map is intended as a general example for a single-core processor for both MPC551x and MPC555x devices. Therefore some memory section names are defined that do not exist for some builds. For example, the Decrementer example does not use the section “intc_hw_branch_table.” Similarly, section “ivor_branch_table” is only used for interrupts on MPC551x, and does not exist for MPC555x applications. NOTE: These examples assume 32 KB of internal SRAM. Some devices, such as MPC5604S, only have 24 KB internal SRAM, so adjustments are needed for link files in that case. Memory Segment Names 0x0000 0000 Memory Section Names boot_flash (64 KB) boot interrupts_flash (64 KB) ivor_branch_table intc_hw_branch_table ivor_handlers contains RCHW, start vector MPC551x core interrupts 0x0001 0000 0x0002 0000 internal_flash (384 KB) intc_sw_isr_vector_table (code, etc.) For INTC HW vector mode. If using this mode, table starts at: 0x4000 0800 (MPC551x) 0x4000 0000 (MPC555x) For INTC SW vector mode. If using this mode, table starts at 0x0002 0000. 0x0007 FFFF 0x4000 0000 internal_ram (31 KB) (data, etc.) stack_ram (1 KB) (stack) 0x4000 7FFF Figure 41. Memory Map for Executing from Internal Flash (IVPR Value is Passed as 0x0001 0000) Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 235 B.2.1 CodeWarrior Link File for Execution from Internal Flash /* 5500_flash.lcf - Simple minimal MPC5500 link file using 32 KB SRAM */ /* Sept 20 2007 SM, DF initial version */ /* May 09 208 SM: Put stack in it's own 1KB (0x400) memory segment */ MEMORY { boot_flash: interrupts_flash: internal_flash: internal_ram: stack_ram: } org org org org org = = = = = 0x00000000, 0x00010000, 0x00020000, 0x40000000, 0x40007C00, len len len len len = = = = = 0x00010000 0x00010000 0x00060000 0x00007C00 0x0400 /* This will ensure the rchw and reset vector are not stripped by the linker */ FORCEACTIVE { "bam_rchw" "bam_resetvector"} SECTIONS { .boot LOAD (0x00000000) : {} > boot_flash /* LOAD (0x0) prevents relocation by ROM copy during startup */ GROUP : { /* Note: _e_ prefix enables load after END of that specified section */ .ivor_branch_table (TEXT) LOAD (ADDR(interrupts_flash)) : {} .intc_hw_branch_table (TEXT) LOAD (_e_ivor_branch_table) ALIGN (0x800) : {} .ivor_handlers (TEXT) LOAD (_e_intc_hw_branch_table) : {} /* Each MPC555x handler require 16B alignmt */ } > interrupts_flash GROUP : { .intc_sw_isr_vector_table ALIGN (2048) : {} /* For INTC in SW Vector Mode */ .text : { *(.text) *(.rodata) *(.ctors) *(.dtors) *(.init) *(.fini) *(.eini) . = (.+15); } .sdata2 : {} extab : {} extabindex : {} } > internal_flash Qorivva Simple Cookbook, Rev. 4 236 Freescale Semiconductor GROUP : { .data (DATA) : .sdata (DATA) : .sbss (BSS) : .bss (BSS) : .PPC.EMB.sdata0 : {} .PPC.EMB.sbss0 : {} } > internal_ram {} {} {} {} } /* Freescale CodeWarrior compiler address designations */ _stack_addr = ADDR(stack_ram)+SIZEOF(stack_ram); _stack_end = ADDR(stack_ram); /* These are not currently being used _heap_addr = ADDR(.bss)+SIZEOF(.bss); _heap_end = ADDR(internal_ram)+SIZEOF(internal_ram); */ __IVPR_VALUE = ADDR(interrupts_flash); /* L2 SRAM Location (used for L2 SRAM initialization) */ L2SRAM_LOCATION = 0x40000000; Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 237 B.2.2 Diab Link File for Internal Flash Execution /* 5500_flash.lin - Simple minimal MPC5500 link file for 512 KB flash, 32 KB SRAM */ /* Jul 05 2007 S.M. Initial version. */ /* May 09 2008 S.M. Put stack in separate memory segment MEMORY { /*******************************************************************/ /* Address Length Use */ /*******************************************************************/ /* 0000_0000-0000_FFFF 64 KB Flash- RCHW & start vector */ /* 0001_0000-0001_FFFF 64 KB Flash- Interrupt area */ /* 0002_0000-0007_FFFF 384 KB Flash- Available for code, etc */ /* 4000_0000-4000_7BFF 31 KB Internal RAM except stack */ /* 4000_7C00-4000_7FFF 1 KB Internal RAM - stack */ /*******************************************************************/ boot_flash: org = 0x00000000, len = 0x10000 interrupts_flash: org = 0x00010000, len = 0x10000 internal_flash: org = 0x00020000, len = 0x60000 internal_ram: org = 0x40000000, len = 0x7C00 stack_ram: org = 0x40007C00, len = 0x0400 } SECTIONS { .boot : {} > boot_flash GROUP : { .ivor_branch_table : {} /* For MPC5516 core interrupts */ .intc_hw_branch_table ALIGN (2048): {} /* For INTC in HW Vector Mode */ .ivor_handlers : {} /* For core interrupt handlers */ } > interrupts_flash GROUP : { .intc_sw_isr_vector_table ALIGN (2048) .text (TEXT) : { *(.text) *(.rodata) *(.ctors) *(.dtors) *(.init) *(.fini) *(.eini) . = (.+15) & ~15; } .sdata2 (TEXT) : {} } > internal_flash : {} /* For INTC SW Vector Mode */ GROUP : { .data (DATA) LOAD(ADDR(.sdata2)+SIZEOF(.sdata2)) : {} .sdata (DATA) LOAD(ADDR(.sdata2)+SIZEOF(.sdata2)+SIZEOF(.data)) : {} .sbss (BSS) : {} .bss (BSS) : {} } > internal_ram Qorivva Simple Cookbook, Rev. 4 238 Freescale Semiconductor } __IVPR_VALUE = ADDR(interrupts_flash); /* Pass address to to load to IVPR */ __SP_INIT = ADDR(stack_ram)+SIZEOF(stack_ram); __SP_END = ADDR(stack_ram); __DATA_ROM = ADDR(.sdata2)+SIZEOF(.sdata2); __DATA_RAM = ADDR(.data); __DATA_END = ADDR(.sdata)+SIZEOF(.sdata); __BSS_START = ADDR(.sbss); __BSS_END = ADDR(.bss)+SIZEOF(.bss); __HEAP_START= ADDR(.bss)+SIZEOF(.bss); __HEAP_END = ADDR(internal_ram)+SIZEOF(internal_ram); Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 239 B.2.3 Green Hills Link File for Execution from Internal Flash /* 5500_flash.ld - Example minimal MPC5500 link file- 512 KB flash, 32 KB SRAM */ /* May 09 2008 S.M., G.L. Initial version. */ MEMORY { /*******************************************************************/ /* Address Length Use */ /*******************************************************************/ /* 0000_0000-0000_FFFF 64 KB Flash- RCHW & start vector */ /* 0001_0000-0001_FFFF 64 KB Flash- Interrupt area */ /* 0002_0000-0007_FFFF 384 KB Flash- Available for code, etc */ /* 4000_0000-4000_7BFF 31 KB Internal RAM except stack */ /* 4000_7C00-4000_7FFF 1 KB Internal RAM - Stack */ /*******************************************************************/ boot_flash : ORIGIN = 0x00000000, LENGTH = 64K interrupts_flash : ORIGIN = . , LENGTH = 64K internal_flash : ORIGIN = . , LENGTH = 384K internal_ram : ORIGIN = 0x40000000, LENGTH = 31K stack_ram : ORIGIN = . , LENGTH = 1K } CONSTANTS { stack_reserve = 1K heap_reserve = 1K } SECTIONS { // FLASH SECTIONS .boot : > boot_flash .ivor_branch_table : > interrupts_flash .intc_hw_branch_table ALIGN(2048) : > . .ivor_handlers : > . .init .text .syscall .rodata .sdata2 .secinfo .fixaddr .fixtype .CROM.PPC.EMB.sdata0 .CROM.sdata .CROM.data /* For MPC5516 core interrupts */ /* For INTC in HW Vector Mode */ /* For core interrupt handlers */ : > internal_flash : : : : : : : > > > > > > > . . . . . . . CROM(.PPC.EMB.sdata0) : /* CROM(.sdata) : /* CROM(.data) : /* /* For GHS hostio support */ /* For GHS startup code */ > . compressed initialized data */ > . compressed initialized data */ > . compressed initialized data */ Qorivva Simple Cookbook, Rev. 4 240 Freescale Semiconductor // RAM SECTIONS .PPC.EMB.sdata0 ABS .PPC.EMB.sbss0 CLEAR ABS .sdabase ALIGN(8) .sdata .sbss .data .bss .heap ALIGN(16) PAD(heap_reserve) : : : : : : : : > > > > > > > > internal_ram . . . . . . . .stack ALIGN(16) PAD(stack_reserve) : > stack_ram /* SP init = addr stack_ram + pad */ __STACK_SIZE __SP_END = stack_reserve; = ADDR(.stack); __IVPR_VALUE = MEMADDR(interrupts_flash); /* Pass address to load to IVPR */ // // These special symbols mark the bounds of RAM and ROM memory. // They are used by the MULTI debugger. // __ghs_ramstart = MEMADDR(internal_ram); __ghs_ramend = MEMENDADDR(stack_ram); __ghs_romstart = MEMADDR(boot_flash); __ghs_romend = MEMENDADDR(internal_flash); // // These special symbols mark the bounds of RAM and ROM images of boot code. // They are used by the GHS startup code (_start and __ghs_ind_crt0). // __ghs_rambootcodestart = 0; __ghs_rambootcodeend = 0; __ghs_rombootcodestart = ADDR(.boot); __ghs_rombootcodeend = ENDADDR(.fixtype); } Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 241 B.3 Additions to Startup File for Flash Executable The startup file, often called the “crt0” file, has two modifications below for a flash executable. RAM executable example programs perform these functions by executing a debugger script file, typically when the debugger starts up. 1. Add a .boot section that contains: — the Reset Configuration Half Word (RCHW) — the vector to the first line of code, typically labelled “_start” 2. Code to initialize all of internal RAM. This must be done before using RAM, to avoid ECC errors. Below is example .boot code that is added to the startup file of these flash examples. The file is then renamed, such as “crt0new.s,” and included in the build. MPC563x devices have a second watchdog that can be enabled in the RCHW. For these examples, watchdog(s) are disabled by the RCHW for flash targets. .section .boot .LONG 0x005A0000 # RCHW: WTE = SWT = PS = VLE = 0, BOOTID = 0x5A .LONG _start Sample code to initialize RAM is listed in the MPC55xx reference manuals. The code below is from the MPC5553/MPC5554 Reference Manual to initialize 64 KB SRAM. It is inserted in the startup file. init_l2RAM: lis r11, 0x4000 # base address of RAM, 64-bit word aligned li r12, 512 # loop counter for 64 bits x 512 = 64 KB RAM mtctr r12 init_l2ram_loop: stmw r0, 0(r11) # write all 32 GPRs to RAM addi r11, r11, 128 # increment the ram pointer; 32 GPRs x 4 B = 128 bdnz init_l2ram_loop # loop for 64 KB of RAM Qorivva Simple Cookbook, Rev. 4 242 Freescale Semiconductor B.4 “Makefile” for Building Flash & RAM Executable Programs # makefile : Sample makefile for MPC5500 to work with Diab v5.5.3.1 # Rev 1 - Jan, 2003 - based on example from S.M. & P.S. # Rev 2 - March, 2006 - updated for DWARF 2.0 output # Rev 3 - Feb, 2007 - modified to provide both RAM and FLASH output files OBJS-RAM= main.o handlers.o ivor_branch_table.o # For MPC551x only OBJS-FLASH= main.o handlers.o crt0new.o ivor_branch_table.o # For MPC551x only #OBJS-RAM= main.o handlers.o # For MPC555x only #OBJS-FLASH= main.o handlers.o crt0new.o # For MPC555x only CC = dcc AS = das LD = dcc DUMP = ddump COPTS = -tPPC5554ES:cross -@E+err.log -g -O -c -Xdebug-dwarf2-extensions-off / -I ..\MPC5500 AOPTS = -tPPC5554ES:cross -@E+err.log -g LOPTS-RAM = -tPPC5554ES:cross -@E+err.log -Ws -m6 -lc -l:crt0.o LOPTS-FLASH = -tPPC5554ES:cross -@E+err.log -Ws -m6 -lc EXECUTABLE-RAM = DEC-ram EXECUTABLE-FLASH = DEC-flash .SUFFIXES: .c .s .c.o : $(CC) $(COPTS) -o $*.o $< .s.o : $(AS) $(AOPTS) $< default: $(EXECUTABLE-RAM).elf $(EXECUTABLE-RAM).s19 $(EXECUTABLE-FLASH).elf $(EXECUTABLE-FLASH).s19 $(EXECUTABLE-RAM).elf: makefile $(OBJS-RAM) $(LD) $(LOPTS-RAM) $(OBJS-RAM) -o $(EXECUTABLE-RAM).elf -Wm 5500_ram.lin / > $(EXECUTABLE-RAM).map $(DUMP) -tv $(EXECUTABLE-RAM).elf >>$(EXECUTABLE-RAM).map # Generate s record file $(EXECUTABLE-RAM).s19: $(EXECUTABLE-RAM).elf $(DUMP) -Rv -o $(EXECUTABLE-RAM).s19 $(EXECUTABLE-RAM).elf $(EXECUTABLE-FLASH).elf: makefile $(OBJS-FLASH) $(LD) $(LOPTS-FLASH) $(OBJS-FLASH) -o $(EXECUTABLE-FLASH).elf / -Wm 5500_flash.lin > $(EXECUTABLE-FLASH).map $(DUMP) -tv $(EXECUTABLE-FLASH).elf >>$(EXECUTABLE-FLASH).map # Generate s record file $(EXECUTABLE-FLASH).s19: $(EXECUTABLE-FLASH).elf $(DUMP) -Rv -o $(EXECUTABLE-FLASH).s19 $(EXECUTABLE-FLASH).elf clean: del *.o Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 243 B.5 VLE Implementation - CodeWarrior Project CodeWarrior’s project wizard allows selection of the project to be VLE. If converting an existing project to use VLE, the following must be done. In the project’s settings, change the following: 1. Target — Build Extras: If “Use External Debugger” is checked, the debugger may need to know code is VLE. For example, the PEMicro debugger would have the Argument: LOADTORAM -RESETFILE “mpc5516vle.mac” 2. Language Settings — C/C++ PreProcessor: Set the preprocessor macro: “#define VLE_IS_ON 1”. This gets used in MPC55xx_init.c when setting up the RCHW. 3. Code Generation — EPPC Processor: In the “e500/Zen Options” box, check the item “Generate VLE Instructions.” The linker files must indicate code sections to be VLE. Example: SECTIONS { .__bam_bootarea LOAD (0x00000000): {} > resetvector GROUP : { .ivor_branch_table_p0 (VLECODE) LOAD (0x00001000) : {} .intc_hw_branch_table_p0 LOAD (0x00001800): {} .__exception_handlers_p0 (VLECODE) LOAD (0x00001100) : {} } > exception_handlers_p0 GROUP : { .intc_sw_isr_vector_table_p0 ALIGN (2048) : {} .init : {} .init_vle (VLECODE) : { *(.init) *(.init_vle) } .text : {} .text_vle (VLECODE) ALIGN(0x1000): { *(.text) *(.text_vle) } .rodata (CONST) : { *(.rdata) *(.rodata) } .ctors : {} .dtors : {} extab : {} extabindex : {} } > internal_flash Any assembly function must either be encapsulated in a C function or use VLE mnemonics, which start with “e_” or “se_”. Qorivva Simple Cookbook, Rev. 4 244 Freescale Semiconductor Appendix C MPC56xxB/P/S Peripheral Clocks Clocks to peripherals are sometimes not connected after reset. If a peripheral does not have a clock, there will be a bus error when attempting to access any of its registers. Software must initialize in two ways: C.1 Peripheral Clock Gating on a Mode Basis Clocks are gated to peripherals based on ME_RUN_PCx, ME_LP_PCx registers and each peripheral’s ME_PCTLx register. This allows peripherals to be clocked in some modes, unclocked in others. The table below lists the ME_PCTLx registers and the peripherals they control. Table 108. Peripheral Control Registers ME_PCTL # Peripheral MPC56xxB MPC56xxP MPC56xxS 4-5 DSPI 0:1 Y Y Y 6 DSPI 2 Y Y — 7 DSPI 3 — Y — 10 QuadSPI — — Y 16 FlexCAN 0 Y Y Y 17 FlexCAN 1 Y — Y 18 FlexCAN 2 Y — — 19 - 21 FlexCAN 3:5 Y — — 23 DMA Mux — — Y 24 FlexRAY — Y — 26 Safety Port — Y — 32 ADC 0 Y Y Y 33 ADC 1 — Y — 35 CTU 0 — Y — 38 - 39 eTimer 0:1 — Y — 41 FlexPWM 0 — Y — 44 I2C DMA 0 Y — Y 45 - 47 I2C DMA 1:3 — — Y 48 - 49 LIN FLEX 0:1 Y Y Y 50 - 51 LIN FLEX 2:3 Y — — 56 Gauge Driver — — Y 57 CTUL Y — — 60 CAN Sampler Y — Y 61 LCD — — Y 62 Sound Gen. — — Y 63 DCU — — Y Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 245 Table 108. Peripheral Control Registers C.2 ME_PCTL # Peripheral MPC56xxB MPC56xxP MPC56xxS 68 SIUL Y — Y 69 WKPU Y — Y 72 - 73 eMIOS 0:1 Y — Y 91 RTC/API Y — Y 92 PIT/RTI Y Y Y 104 CMU Y — Y Clock Generation on a Peripheral or Peripheral Set Basis In addition to the peripheral clock gating, some peripherals also require generating a clock to them. This clock generation is done on a “peripheral set” basis where all peripherals in a group run at the same frequency. Clocks to those peripherals may be disabled after reset and must be enabled with a clock divider value. Sometimes you must specify the clock source such as PLL, external oscillator, etc. The following table lists the sets and the registers that must be configured to use any peripheral in that set. Table 109. Peripheral Set Clock Generation Registers MPC56xxB Peripheral Set Peripherals MPC56xxP (after cut 1) Registers to Enable and Generate Clock Peripherals Registers to Enable and Generate Clock MPC56xxS (after cut 1) Peripherals Registers to Enable and Generate Clock 1 LINFlex I2C CGM_SC_DC0 Motor Control CGM_AC0_SC LINFlex CGM_SC_DC0 CGM_AC0_DC0 I2C Motor Control Stall Detect Sound Gen. LCD 2 FlexCAN DSPI CGM_SC_DC1 CMU Monitor CGM_AC1_SC FlexCAN CGM_SC_DC1 CGM_AC1_DC0 CAN Sampler DSPI 3 eMIOS ADC CTU CGM_SC_DC2 Safety Port CGM_AC2_SC ADC CGM_AC2_DC0 CGM_SC_DC2 FlexRay CGM_AC3_SC DCU CGM_AC3_DC0 CGM_AC0_SC — — eMIOS_0 CGM_AC1_SC CGM_AC1_DC0 — eMIOS_1 CGM_AC2_SC CGM_AC2_DC0 — QuadSPI CGM_AC3_SC Qorivva Simple Cookbook, Rev. 4 246 Freescale Semiconductor Qorivva Simple Cookbook, Rev. 4 Freescale Semiconductor 247 How to Reach Us: Information in this document is provided solely to enable system and software Home Page: freescale.com implementers to use Freescale products. There are no express or implied copyright Web Support: freescale.com/support information in this document. licenses granted hereunder to design or fabricate any integrated circuits based on the Freescale reserves the right to make changes without further notice to any products herein. Freescale makes no warranty, representation, or guarantee regarding the suitability of its products for any particular purpose, nor does Freescale assume any liability arising out of the application or use of any product or circuit, and specifically disclaims any and all liability, including without limitation consequential or incidental damages. “Typical” parameters that may be provided in Freescale data sheets and/or specifications can and do vary in different applications, and actual performance may vary over time. All operating parameters, including “typicals,” must be validated for each customer application by customer’s technical experts. Freescale does not convey any license under its patent rights nor the rights of others. Freescale sells products pursuant to standard terms and conditions of sale, which can be found at the following address: http://www.reg.net/v2/webservices/Freescale/Docs/TermsandConditions.htm Freescale, the Freescale logo, AltiVec, C-5, CodeTest, CodeWarrior, ColdFire, C-Ware, Energy Efficient Solutions logo, Kinetis, mobileGT, PowerQUICC, Processor Expert, QorIQ, Qorivva, StarCore, Symphony, and VortiQa are trademarks of Freescale Semiconductor, Inc., Reg. U.S. Pat. & Tm. Off. Airfast, BeeKit, BeeStack, ColdFire+, CoreNet, Flexis, MagniV, MXC, Platform in a Package, QorIQ Qonverge, QUICC Engine, Ready Play, SafeAssure, SMARTMOS, TurboLink, Vybrid, and Xtrinsic are trademarks of Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. The Power Architecture and Power.org word marks and the Power and Power.org logos and related marks are trademarks and service marks licensed by Power.org. © 2005–2010 Freescale Semiconductor, Inc. Document Number: AN2865 Rev. 4 04/2010